Abstract Book

Table of Contents

Talks ...... 1 Posters ...... 69

Student presentations are indicated with (S) PopGroup52 Oxford 2019 Talks

TALKS

The mystery of the U-shaped spectra

F. Freund, S. Matuszewski, E. Kerdoncuff, J. Jensen, M. Lapierre, A. Lambert and G. Achaz

Collège de France, France

Since the advent of the neutral theory of molecular evolution the standard neutral model (mutation-drift equilibrium) has been elevated as a reference model in light of which we routinely interpret population genetics data. Using either forward diffusion or backward coalescent approaches, it can be shown that the expected distribution of mutation frequencies (the so-called sites frequency spectrum, SFS) is proportional to 1/f, where f is the frequency of the mutations. We have collected a large collection of genome-wide SFS (averages of thousands of loci) from a diverse set of organisms, both eukaryotic (plants, fungi, animals) and prokaryotic (bacteria and archea). All the observed SFS were compared to the standard neutral SFS and, with only few exceptions, none fits the expectations of the reference model. The observed SFSs typically show an excess of low and high frequency variants, leading to what is known as U-shaped spectra. Including simple demographies (e.g. monotonic growth/decline) does not result in a perfect fit, neither can the inclusion of mis-orientation errors (erroneous swaps of derived and ancestral frequencies). We have therefore derived theoretical expectations from multiple mergers coalescent (MMC) models (i.e. Beta-coalescents and Xi-coalescents, that both can tuned by a single parameter) together with exponential growth and mis-orientation. MMC models emerge from modes of evolution where few individual can concentrate the parenthood of a large fraction of the population. This include some selective regimes (e.g. the genetic draft) or extremely skewed offspring distribution (e.g. sweep-take reproduction mode). Using both a likely-based approaches and least-square regressions, we show that this three parameter models (mis-orientation, growth rate plus concentrated parenthood) show an excellent fit to many of the observed SFS. We thus discuss the possible replacement of the old reference model by a new one that do fit most of the data (what is expected from a null model !). We further discuss the potential causes of the excellent fit of this new model together the challenging observation that the standard neutral model typically fails to fit the data.

1 PopGroup52 Oxford 2019 Talks

A case of mitochondrial DNA recombination?

Einar Árnason and Katrín Halldórsdóttir

Department of Life and Environmental Sciences, University of Iceland, Iceland

It is generally accepted that mitochondrial DNA does not recombine although several recent examples challenge this dogma. In a study of codfish genomics we have come across a case of potential recombination of widely separated mitochondrial DNAs. Is it real or an artifact of the method? We discuss the evidence for mitochondrial differences among the taxa in light of the whole genome nuclear differences.

Tracing the source of Gastroenteritis with Machine Learning (S)

Nicolas Arning, Daniel J. Wilson

Big Data Institute, University of Oxford, UK

Gastroenteritis is a food-borne disease commonly caused by the bacterium Campylobacter jejuni. Human infection sources can be under-cooked meat, contact with animal faeces or environmental sources like contaminated drinking water. Attributing the source of a gastroenteritis outbreak is important for applying public health regulations. Previous efforts depend on statistical inference through the comparison of genetic sequence between human and source samples. Our aim is to develop a new Machine Learning based application for source attribution to outperform the existing methods and broaden the spectrum of viable input. The choice of Machine Learning as a solution is due to its underuse, albeit reported success, in genomic contexts. This aim was implemented through unbiased choice of tools and use of a k-mer based approach applicable to any form of genetic sequence. This is especially useful with the ever growing amount of whole genome data available. We report a slight increase in accuracy and present a protocol which can be used for input forms unavailable for use with previous methods. Our results confirms the applicability of Machine learning in genomic contexts and can be used as a stepping stone for similar future endeavours.

2 PopGroup52 Oxford 2019 Talks

Investigating Hybrid Male Sterility in the House Mouse (S)

Paigan Aspinall, Shauni Doms, John Baines, Leslie Turner

University of Bath, UK

Mechanisms of the early stages of speciation can be understood by identifying barriers to gene flow between naturally hybridising species. A common barrier between recently diverged species is reduced fertility in hybrids, particularly males. The house mouse subspecies - Mus musculus musculus and Mus musculus domesticus - meet and interbreed in a natural hybrid zone running through central Europe. Hybrid male sterility has been observed as a barrier to gene flow in these house mouse subspecies. Previous studies of hybrid male infertility have focussed on gross phenotyping, which is likely to be a result of multiple genetic incompatibilities. Here we use crosses of strains derived from wild-caught M. m musculus-M. m. domesticus hybrids to investigate reproductive phenotypes in order to determine the number, type, severity and prevalence of fertility defects present. Unlike much of the previous hybrid male sterility work, we perform detailed histological analysis to identify the finer defects that are causing gross phenotypes. Complete sterility appears to be absent in hybrid crosses, but a large proportion of males exhibit fertility phenotypes below the range found in pure subspecies, and as a result likely suffer reduced fertility. The comparison of phenotypes indicates reduced hybrid fertility is highly variable among individuals, suggesting that multiple underlying genetic incompatibilities are responsible for segregation in the hybrid zone. These results are being used in quantitative trait locus (QTL) mapping to identify loci contributing to sterility phenotypes.

3 PopGroup52 Oxford 2019 Talks

Epistatic interactions and the evolution of self-fertilisation

Diala Abu Awad, Denis Roze

Technical University of Munich, Germany

Self-fertilization is a wide-spread reproductive strategy that is found in both hermaphroditic animals and plants. Natural populations exhibit a large range of self-fertilizing frequencies, from complete outcrossing (no self-fertilization) to almost exclusive self-fertilization. For self-fertilization to evolve, the levels of inbreeding depression (due to segregating deleterious mutations) maintained in populations must be sufficiently low and the purging of deleterious mutations must be possible. If these conditions are met, existing models predict that populations will evolve towards strict self-fertilization, implying that intermediate self-fertilization rates are not a stable state. This, however, does not concur with the existence of apparently stable mixed-mating strategies. This discrepancy between empirical observations and theoretical expectations could be due to the simplistic assumptions made concerning the underlying genetic architecture of traits affecting self- fertilization and fitness, often relying on single locus approximations. Recent works have shown that the consequences of self-fertilization on inbreeding depression in the presence of epistatic fitness interactions can differ greatly from single-locus approximations. It is therefore primordial that more complex genetic architectures are considered when studying the evolution of self-fertilization. Here we present a model based on Fisher's geometric model in which epistatic fitness interactions between loci arise naturally. Self- fertilization is allowed to evolve through mutations at a number of several interacting loci and fitness depends on n quantitative traits under selection. We determine the conditions necessary for the evolution of self-fertilization and whether or not mixed-mating strategies are stable in this context.

Holobiont Suture Zones

Joëlle Goüy de Bellocq, Wasimuddin, Alexis Ribas, Josef Bryja, Jaroslav Piálek, Stuart J.E. Baird

Czech Academy of Sciences, Prague

Parasite hybrid zones resulting from host secondary contact have never been described in nature although parasite hybridization is well known and secondary contact should affect them similarly to free-living organisms. When host populations are isolated, diverge, and re-contact, intimate parasites (host specific, direct life cycle) carried during isolation will also meet and so may form parasite hybrid zones. If so, we hypothesize these should be narrower than the host’s hybrid zone as shorter parasite generation time allows potentially higher divergence. We investigate multi-locus genetics of two parasites across the European house mouse hybrid zone. We find each host taxon harbours its own parasite taxa. These also hybridize: parasite hybrid zones are significantly narrower than the host’s. Here we show a host hybrid zone is a suture zone for a subset of its parasite community and highlight the potential of such systems as windows on the evolutionary processes of host-parasite interactions and recombinant pathogen emergence.

4 PopGroup52 Oxford 2019 Talks

Inferring the selective pressures acting on insertions and deletions in the great tit genome (S)

Henry J Barton, Kai Zeng

The University of Sheffield, UK

Insertions and deletions (INDELs) remain understudied, despite being the most common form of genetic variation after single nucleotide polymorphisms. This stems partly from the challenge of correctly identifying the ancestral state of an INDEL and thus identifying it as an insertion or a deletion. Erroneously assigned ancestral states can skew the site frequency spectrum, leading to artificial signals of selection. Consequently, the selective pressures acting on INDELs are at present poorly resolved. To tackle this issue, we have recently published a maximum likelihood approach to estimate the mutation rate and the distribution of fitness effects (DFE) for insertions and deletions. Our approach estimates and controls for the rate of ancestral state miss-identification, overcoming issues plaguing previous INDEL studies. Here we apply the method to INDEL polymorphism data from 10 high coverage (44X) European great tit (Parus major) genomes. We demonstrate coding INDELs are under strong purifying selection with a small proportion making it into the population (~5%). However, among fixed coding INDELs, 71% of insertions and 86% of deletions are fixed by positive selection. In non-coding regions we see evidence of linked selection reducing INDEL diversity below background levels, both in proximity to exons and in areas of low recombination.

Selection at linked sites in regions of low crossing over

Hannes Becher, Benjamin Jackson, Brian Charlesworth

University of Edinburgh, UK

Selection acting on linked sites has profound impacts on variation and evolution: it causes the genomic patterns characteristic of selective sweeps, it reduces the efficacy of selection when acting on multiple loci (Hill-Robertson interference), and it reduces genetic diversity (by more than 15% in Drosophila melanogaster). The lower the recombination rate in a genomic regions, the greater the effect of selection at linked sites. Using whole genome sequencing data from two African populations of Drosophila simulans, we explore the interplay between the rate of crossing over and selection at linked sites, with a special focus on non-crossover genomic regions. Both populations show signs of recent population expansion, which affects the action of selection at linked sites on a complex fashion. On autosomes, this causes a positive relation between the skew towards rare variants and the recombination rate. Our findings agree with the results of simulations carried out in SLiM2.

5 PopGroup52 Oxford 2019 Talks

Exaggerated heterochiasmy in a sexually dimorphic fish (Poecilia reticulata)

Roberta Bergero, Deborah Charlesworth

University of Edinburgh, UK

Sexual dimorphism (heterochiasmy) of meiotic recombination is known from several kinds of organisms, being extreme in Diptera, where neither pairing, nor crossing over or gene conversion occurs in male meiosis. When one sex has no or very infrequent recombination, it is usually the heterogametic sex, as in Lepidoptera, with achiasmate females. In studying sex linkage in the guppy (Poecilia reticulata), we uncovered a case of sexual dimorphism in recombination patterns. Genetic mapping shows that meiotic recombination occurs freely during female meiosis, whereas in male meiosis crossovers occur only in small regions at the chromosome tips for all chromosomes tested, including chromosome 12, which carries the sex determining locus. Genome resequencing of males and females also provides evidence for rare crossing over events across most of the sex chromosomes, with many SNPs revealing distinctive Y chromosome haplotypes. Since highly localized recombination in male meiosis is genome-wide in the guppy, it might be ancestral. However, genetic map information from related fish (a cross between Xiphphorus species, and an F2 in medaka), suggest that heterochiasmy probably evolved in the guppy lineage. By virtue of the guppy’s heterochiasmy, mutations arising on the Y chromosome are expected to be transmitted mainly to the male progeny. This should facilitate the invasion and maintenance of male-beneficial, but female-detrimental mutations since females rarely experience the deleterious effect of these mutations. The sexually dimorphic recombination pattern can thus potentially explain the observed enrichment of male coloration genes on the guppy Y chromosome.

How do new mating types arise in cooperative non-self recognition system of flowering plants?

Katarina Bodova, Tadeas Priklopil, David Field, Nick Barton, Melinda Pickup

Comenius University, Bratislava, Slovakia

Self-incompatibility is widespread in flowering plants and functions to prevent self- fertilization and the deleterious effects of inbreeding depression. The incredible diversity observed in natural populations is maintained by negative frequency dependent selection, a mechanism where a rare allele has a selective advantage. Yet one of the biggest and most intriguing questions in the evolution of SI is how new alleles evolve. Here we investigate three main questions in a non-self-recognition system of self-incompatibility. How can SI evolve from an initial system of self-compatible individuals? How can an established SI system diversify to gain a new S-allele? Are the processes of establishment and diversification of S-alleles consistent with reality in terms of the observed number of S- alleles?

6 PopGroup52 Oxford 2019 Talks

Investigating the post-embryonic role and regulation of Ultrabithorax (S)

Alexandra Buffry, Sebastian Kittelmann, Franziska Anni Franke, Korneel Hens, Saad Arif and Alistair P. McGregor

Oxford Brookes University, UK

Hox genes play fundamental roles in the patterning of animal body plans. Early expression of Hox genes in different domains along the embryonic anterior-posterior axis are responsible for determining segmental identity. However, Hox genes also function post- embryonically to specify fine scale morphology such as the size, shape and appearance of limbs and other organs in Drosophila. Less is known about these post-embryonic roles and their regulation but, given that changes in Hox gene expression is often correlated with the phenotypic evolution, it is essential that we investigate the spatial regulation of Hox genes at all developmental stages to fully understand their role in conferring distinct and diverse patterns. In Drosophila, Ultrabithorax (Ubx) regulates the morphology of the T2 and T3 legs and has been implicated in variation in leg trichome patterning. We have investigated this role further by generating ATAC-seq data for T2 legs to identify a novel enhancer of Ubx that contributes to trichome patterning. Dissection of this putative regulatory region reveals that postembryonic regulation of Ubx is highly complex. To determine how this expression of Ubx is regulated we have carried out a yeast one hybrid screen on this novel enhancer to identify putative transcription factors. Our results will help to reveal how Hox genes are wired into the postembryonic gene regulatory networks responsible for fine-scale patterning of animal bodies.

Possibilities for population genetic control with Y-linked editors

Austin Burt, Anne Deredec

Imperial College, UK

Some pest species continue to cause considerable harm, and new methods of control are needed. Recent advances in molecular biology raise the possibility of new genetic interventions for population control. In principle, efficient control strategies can be developed to exploit the fact that the Y chromosome in male-heterogametic species is transmitted solely from fathers to sons, and is selected for only based on its impacts on male fitness. Simple population genetic and population dynamic models are used to illustrate how Y-linked editors that target the fitness of female descendants could be used to efficiently control pest populations.

7 PopGroup52 Oxford 2019 Talks

Mito-nuclear genotypes modulate metabolism with downstream fitness consequences

M.F. Camus, V.N. Kotiadis, H. Carter, N. Lane

University College London, UK

Mitochondria are essential for eukaryote evolution, taking centre-stage in the process of cellular respiration. This process is regulated via a series of finely coordinated interactions encoded by two obligate genomes – nuclear and mitochondrial. Both genomes are required for the production of cellular energy, and thus their harmonious interaction is essential to maintain mitochondrial integrity and the viability of eukaryote life. Although these interactions are highly important for organismal fitness, we do not know the mechanisms that underpin these processes. Here we aim to unravel these complex inter- genomic interactions by examining the physiological and fitness consequences of mito- nuclear incompatibilities using Drosophila melanogaster. We additionally examine how mito-nuclear interactions impact fitness across different levels of cellular stress (induced by antioxidants). We find that antioxidant stress has a significant effect on all fly lines, however the magnitude of the response was heavily dependent on the mito-nuclear combination of the given fly strain. We then hone in the physiological responses underpinning the observed fitness outcomes, and find that metabolic processes also greatly depend on the mito-nuclear genotype. By examining physiological traits, we create more direct link between genotype, metabolic phenotype and fitness. These findings highlight the need to examine intermediate phenotypes in order to fully comprehend fitness consequences of diverse genotypes.

Predicting evolutionary pathways to antimicrobial resistance: A computational biophysics approach

Martin Carballo-Pacheco, Vaitea Opuu, Arjan de Visser, Thomas Simonson, Bartlomiej Waclaw, Rosalind Allen

University of Edinburgh, UK

Microbes are evolving resistance to drugs rendering them inefficient. De novo evolution of resistance is caused by mutations in genes that lead to drug-resistant phenotypes. Evolution of resistance can be understood as trajectories in a fitness landscape, which map a phenotypic effect (resistance to drug) onto the highly multidimensional genotypic space. Features from these landscapes determine how evolution ensues and which phenotypes are accessible in an specific environment. However, measuring fitness landscapes experimentally is challenging due to their high dimensionality. Here, we propose a computational model to predict the binding of antimicrobial drugs to their target protein and build the fitness landscape for antimicrobial resistance by taking into account the atomistic details of both the target and the drug.

8 PopGroup52 Oxford 2019 Talks

Widespread tRNA Deamination In A 900 Million Year Old Premetazoan

Martin Carr

Department of Biological Sciences, University of Huddersfield, UK

Patterns of codon usage bias vary considerably across Metazoa, with both mutation pressure and natural selection contributing to bias. The highly variable nature of their codon usage prevents the reconstruction of ancestral traits using extant metazoan species. Traits conserved across the unicellular lineages related to Metazoa, however, do point to ancestral characteristics in the earliest members of the metazoan lineage (premetazoans). Presented here are patterns of codon usage bias in two choanoflagellates and a filasterean and highlight the high level of conservation across one billion years of evolution. The conserved traits can be considered ancestral to metazoans. In particular, it is shown that codon usage bias is directed towards a set of translationally optimal GC-ending codons. Selection appears to operate through translational efficiency and translational accuracy, with even the most weakly biased genes showing the signature of selection. The major tRNA genes for twofold degenerate amino acids match optimal codons, however this is not the case for three-fold to six-fold amino acids. For these amino acids optimal codons show cytosine at the synonymous position, whereas tRNA anticodons have adenosine at the wobble site. It can be seen that tRNA molecules undergo deamination of adenosine at the wobble site to inosine, facilitating complementary binding to cytosine in optimal codons. Highly biased genes in the three species, and therefore in ancient premetazoans, preferentially use deaminated tRNA molecules in preference to unmodified molecules in order to optimise protein synthesis

Evidence for associative overdominance in low recombination regions of the Drosophila genome

Brian Charlesworth and Hannes Becher

Institute of Evolutionary Biology, University of Edinburgh, UK

Recessive or partially recessive deleterious mutations can retard the rate of loss of variability by genetic drift at linked neutral sites, provided that the product of effective population size and the selection coefficient for the mutations is of order one (associative overdominance, or AOD). It is commonly thought that AOD can only occur in very small populations. However, in regions of the genome with low rates of recombination, selective sweeps and background selection can reduce the effective population size to a fraction of the genome-wide average, suggesting that AOD could be induced by sufficiently weakly selected deleterious mutations. Computer simulations show that this is likely for realistic selection and mutation parameters, for a model of low recombination regions of the Drosophila genome. The results are consistent with observations on the degree of distortion of the site frequency spectra at synonymous sites in these genomic regions in D. melanogaster and D. simulans.

9 PopGroup52 Oxford 2019 Talks

Associations between the sex-determining locus and sex-linked sequence variants in the guppy

Deborah Charlesworth and Roberta Bergero

University of Edinburgh, UK

The guppy (Poecilia reticulata) has a marked enrichment of polymorphic male coloration genes on the Y chromosome, some of which appear completely sex linked, and others partially sex linked. There is evidence that these factors effects are sexually antagonistic (SA), although complete sex linkage will prevent transmission to females, and the partially sex-linked factors are not expressed in females. Nevertheless, it has been proposed that there is ongoing selection in guppy populations causing closer linkage between partially sex-linked SA polymorphisms and the sex-determining locus. The finding of crossover localisation in male meiosis (Bergero & Charlesworth talk) casts doubt on the view that recombination suppression is ongoing. However, differentiation between the sexes is prominent in two regions of the sex chromosome pair (guppy chromosome 12). These regions are located far apart on the chromosome, and the associations were detected in several population samples, indicating that linkage disequilibrium is maintained between the sex-determining locus and other Y-linked variants could be the predicted footprint of the presence of SA polymorphisms.

Natural variation in telomere repeats, a major plant satellite DNA, correlates with flowering time variation

Jae Young Choi, Simon Groen, Michael Purugganan

New York University, USA

Satellite DNAs are tandem repeats of short sequences. The evolution of satellites has been a long-standing interest; however, due to their low complexity, studies have been limited to comparative genomics, while the abundance and evolution of repeats within species remain unexplored. Further, the phenotypic consequences arising from these satellites are largely unknown, begging the question whether these repeats are “junk” DNA or have functional effects to the harboring organism. Here, we examined the simple-repeat library within individuals from the model plant eudicot Arabidopsis thaliana and monocot Oryza sativa to characterize the evolution of plant satellite DNAs. Satellites were characterized from sequencing reads generated from whole genome sequencing 483 A. thaliana and 897 O. sativa individuals. We found that population relationships based on satellites were incongruent with those generated from nucleotide polymorphisms, suggesting independent evolutionary histories between genetic variation and repeat variation. Compared to A. thaliana, O. sativa had three-fold higher satellite repeat content, which was consistent with the three-fold difference in their genome sizes. But the top most abundant satellite repeats were largely concordant between the two plants and few species-specific repeats were observed. The most abundant repeat sequence in both species was the telomere repeat sequence (TRS) “AAACCCT”, where there were 10~20 fold variation between individuals. Most importantly, we found that the abundance of TRS was negatively correlated with flowering time in both plant species. This suggests that in addition to genetic variation, genomic features, such as telomere length, may also directly influence crucial plant life-history traits.

10 PopGroup52 Oxford 2019 Talks

Phylotranscriptomic insights into the diversification of endothermic Thunnus tunas

Adam Ciezarek, Owen Osborne, Oliver N. Shipley, Edward J. Brooks, Sean Tracey, Jaime McAllister, Luke Gardner, Michael J. E. Sternberg, Barbara Block, Vincent Savolainen

Earlham Institute, Imperial College London, UK

Birds, mammals, and certain fishes, including tunas, opahs and lamnid sharks, are endothermic, conserving internally generated, metabolic heat to maintain body or tissue temperatures above that of the environment. Bluefin tunas are commercially important fishes worldwide and some populations are threatened. They are renowned for their endothermy, maintaining elevated temperatures of the oxidative locomotor muscle, viscera, brain and eyes, and occupying cold, productive high-latitude waters. Less cold- tolerant tunas, such as yellowfin tuna, by contrast, remain in warm-temperate to tropical waters year-round, reproducing more rapidly than most temperate bluefin tuna populations, providing resiliency in the face of large scale industrial fisheries. Despite the importance of these traits to not only fisheries, but responses to habitat utilisation and climate change, little is known of the genetic processes underlying the diversification of tunas. In collecting and analysing sequence data across 29,556 genes, we found that parallel selection on standing genetic variation is associated with the evolution of endothermy in bluefin tunas. This includes two shared substitutions in genes encoding glycerol-3 phosphate dehydrogenase, an enzyme that contributes to thermogenesis in bumblebees and mammals, as well as four genes involved in the Krebs cycle, oxidative phosphorylation, β-oxidation and superoxide removal. Using phylogenetic techniques, we further illustrate that the eight Thunnus species are genetically distinct, but found evidence of mitochondrial genome introgression across two species. Phylogeny-based metrics highlight conservation needs for some of these species.

11 PopGroup52 Oxford 2019 Talks

Winter is coming: Haemolymph profiling of queen life-cycle stages reveals key insights into bumblebee diapause

Joe Colgan1, Sive Finlay2, Mark J.F. Brown3; and James C. Carolan4

1School of Biological, Earth and Environmental Sciences, University College Cork, Cork, Ireland 2School of Natural Sciences, Trinity College Dublin, Dublin, Ireland 3School of Biological Sciences, Royal Holloway University of London, Egham, UK 4Department of Biology, Maynooth University, Maynooth, Ireland

Diapause represents an alternative life-history event generally characterised by arrested development, suppressed metabolism and increased stress tolerance. This state allows an organism to avoid prolonged periods of harsh and inhospitable environmental conditions. Despite the fitness benefits of diapause, it can also represent a potential survival bottleneck. This is particularly the case for the bumblebee, Bombus terrestris, whereby reproductive females (queens), enter diapause during the winter months within temperate zones. Despite diapause representing a fundamental period of the queen life-cycle, our understanding of the molecular mechanisms involved in its regulation and maintenance is limited. To address this gap in our understanding, we performed quantitative mass spectrometry-based proteomics on bumblebee queen haemolymph at six time-points throughout the queen life-cycle. Our findings highlight dramatic changes in the protein profile of queen haemolymph before, during and after diapause affecting important biological processes, including immunity, olfaction and reproduction. Our results provide clear molecular evidence for the consequences and benefits of mating at the immune level, benefits that are sustained throughout diapause. In addition, we highlight the potential role of structural proteins in stress tolerance during diapause. Collectively, our results provide a novel insight into the molecular mechanisms by which bumblebees prepare for, survive and recover from diapause, insights that may have implications for our general understanding of these processes in other insect groups.

12 PopGroup52 Oxford 2019 Talks

An exploration of cold tolerance mechanisms across the Drosophila genus

Cook N., Parker D.J., Smith T.K., Ritchie M.G.

University of St Andrews, UK

The ability to maintain function with the onset of seasonal cold has major fitness implications for many organisms and is related to both their current geographic distribution and evolutionary history. The physiological adaptations associated with cold tolerance are well known, particularly in insects where cold acclimation (CA) is used. However, less is known regarding the genetic basis of CA, nor how CA mechanisms vary from an evolutionary perspective. The Drosophila genus, of which the phylogenetic history of many species is well-known, is ideal to examine both phenotypic variation and its underlying mechanism due to its broad ecological range and presence across a range of climatically diverse environments. Previous transcriptomic and RNAi work with D. montana and D. virilis (two very cold- tolerant species) confirmed the role of the Inos gene in CA. Inos encodes the enzyme myo-inositol-1-phosphate synthase which is involved in the biosynthesis of myo-inositol; a known major metabolite in overwintering D. montana. Here, we experimentally determined cold resistance across 9 Drosophila species within a phylogenetic framework and from a range of geographic locations. We are using RNA-seq and GC-MS technologies to generate gene expression and metabolite profiles respectively of cold-acclimated versus non-cold acclimated flies. This will allow us to determine the evolutionary consistency of cold-induced changes in gene expression and metabolism, including Inos and to explore the CA phenotype in more detail from an evolutionary perspective.

13 PopGroup52 Oxford 2019 Talks

Towards population genetics analyses in Chlamydomonas reinhardtii: Patterns of continental and local population structure in field isolates of the model green alga (S)

Rory J. Craig, Rob W. Ness, Katharina B. Böndel, Kazuhara Arakawa, Takashi Nakada, Takuro Ito, Graham Bell, Nick Colegrave & Peter D. Keightley

University of Edinburgh, UK, University of Toronto, Canada

There has been longstanding debate concerning the nature of biogeography and speciation in microbial eukaryotes. Originating from the everything is everywhere paradigm, species were widely considered to be ubiquitous, with unlimited gene flow and extremely large population sizes resulting in low rates of speciation. Conversely, recent taxonomic evidence has shown that many species are endemic, and genetic studies across a wide range of taxa have frequently revealed allopatric population structure, suggesting that the evolutionary histories of microbial species may be more similar to that of higher plants and animals. However, microbial eukaryotes remain severely understudied, and population genomics approaches have seldom been applied to this question. Despite its extensive use as a model organism, very little is known about the biogeography and evolutionary history of Chlamydomonas reinhardtii. We have here analysed whole- genome re-sequencing data for all 42 available field isolates of the species, sampled at nine sites across N. America and Japan. We report evidence for allopatric divergence between continents, and putative allopatric divergence between two highly differentiated lineages within N. America. There is little evidence for genetic structure between isolates within the most densely sampled lineage (Quebec/New England, N=26), where genetic diversity between isolates from one field is approximately equal to that of the entire region. However, we find evidence for substantial between-lineage gene flow, and report that within-lineage many isolates share large haplotype blocks indicative of close relatedness. We delineate the Quebec/New England isolates as suitable for population genetics analyses, and summarise genetic diversity between these isolates.

14 PopGroup52 Oxford 2019 Talks

Temperate grass allergy season defined by spatio-temporal shifts in airborne pollen communities

Georgina L. Brennan1, Caitlin Potter2, Natasha de Vere2,3, Gareth W. Griffith2, Carsten A. Skjøth4, Nicholas J. Osborne5,6, Benedict W. Wheeler5, Rachel N. McInnes7, Yolanda Clewlow7, Adam Barber7, Helen M. Hanlon7, Matthew Hegarty2, Laura Jones1,3, Alexander Kurganskiy4, Francis M. Rowney5, Charlotte Armitage8, Beverley Adams-Groom4, Col R. Ford3, Geoff M. Petch4, The PollerGEN Consortium, and Simon Creer1.

1Molecular Ecology and Fisheries Genetics Laboratory, School of Natural Sciences, Bangor University, Bangor, Gwynedd, Wales, UK 2Aberystwyth University, Aberystwyth, Ceredigion, Wales, UK 3National Botanic Garden of Wales, Llanarthne, Carmarthenshire, Wales, UK 4University of Worcester, Worcester, Worcestershire, UK 5University of Exeter, Truro, Cornwall, UK 6University of New South Wales, Sydney, New South Wales, Australia 7Met Office, Exeter, Devon, UK 8The Woodland Trust, Kempton Way, Grantham, Lincolnshire, UK. 9College of Life and Environmental Science, University of Exeter, Exeter, UK 10Ecological Sciences, The James Hutton Institute, Dundee, Scotland, UK 11The David Hide Asthma & Allergy Research Centre, St Mary's Hospital, Newport, Isle of Wight, UK 12Centre for Ecology & Hydrology, Crowmarsh Gifford, Wallingford, Oxfordshire, UK. 13Environment Department, University of York, Heslington, York, UK

Grass pollen is the world’s most harmful outdoor aeroallergen, affecting up to 400 million people. Sensitivity varies between species and different species of grass flower at different times, but it is not known how airborne communities of grass pollen change in time and space. The persistence and high mobility of grass pollen could result in increasing species richness of airborne pollen over the allergy season. Conversely, if grass pollen does not persist for an extended time in the air, pollen assemblages would be expected to change over the summer months. Here, using targeted and environmental shotgun high throughput sequencing, we tracked the seasonal progression of airborne Poaceae pollen biodiversity across Britain, throughout the grass allergy season. All grass genera displayed discrete, temporally restricted peaks of pollen incidence which varied with latitude, revealing that the taxonomic composition of grass pollen exposure changes substantially across the allergy season. We also found that local flowering events with appropriate temporal delays could be useful for predicting the incidence of particular species of grass pollen in the air. We demonstrate that the community composition of the temperate grass pollen allergy season is not uniform and predict that our findings will facilitate the exploration of links between taxon-specific exposure of harmful grass pollen and disease, with concomitant socio-economic benefits.

15 PopGroup52 Oxford 2019 Talks

Evolution of sexual reproduction: short term consequences of the Hill-Robertson effect

Daniel J.M. Crouch

Wellcome Centre for Human Genetics, University of Oxford. UK

The high prevalence of sexual reproduction remains mysterious, as it poses clear evolutionary drawbacks compared to reproducing asexually. Several possible explanations exist, with one of the most likely being that finite population size causes linkage disequilibria to randomly generate and impede the progress of natural selection. Deleterious linkage disequilibria accumulate disproportionally relative to advantageous linkage disequilibria, given sufficient time (Hill-Robertson effect), thereby giving an advantage to breaking up linkage disequilibrium via sexual recombination. However, it is not well understood how this process works over the time scale of a few generations. For an arbitrary number of loci, the expected change in mean fitness between consecutive generations, a measure of short-term evolutionary success, is shown under random environmental influences to be related to the autocovariance in mean fitness between the generations, capturing the effects of stochastic forces such as genetic drift. Interaction between genetic drift and natural selection, due to randomly generated linkage disequilibria, is demonstrated to be one possible source of mean fitness autocovariance. This suggests a possible role for sexual reproduction in reducing the negative effects of genetic drift, thereby improving the short-term efficacy of natural selection.

Widespread gene duplication and adaptive evolution in the RNA interference pathways of the Drosophila obscura group

Danang Crysnanto, Darren J. Obbard

University of Edinburgh, UK

RNA interference (RNAi) related pathways provide defense against viruses and transposable elements, and have been implicated in the suppression of meiotic drive elements. Genes in these pathways often exhibit high levels of adaptive substitution, and over longer timescales show frequent gene duplication and loss—most likely as a consequence of their role in mediating conflict with these parasites. This is particularly striking for Argonaute 2 (Ago2), which is ancestrally the key effector of antiviral RNAi in insects, but has repeatedly formed new testis-specific duplicates in the recent history of the Drosophila obscura group. Here we take advantage of publicly available genomic and transcriptomic data to identify six further RNAi-pathway genes that have duplicated in this clade of Drosophila, and examine their evolutionary history. As seen for Ago2, we observe high levels of adaptive amino-acid substitution and changes in sex-biased expression in many of the paralogs. However, our phylogenetic analysis suggests that co-duplications of the RNAi machinery were not synchronous, and our expression analysis fails to identify consistent male-specific expression. These results confirm that RNAi genes, including genes of the antiviral and piRNA pathways, undergo frequent independent duplications and that their history has been particularly labile within the Drosophila obscura group. However, they also suggest that the selective pressures driving these changes have not been consistent, implying that more than one selective agent may be responsible.

16 PopGroup52 Oxford 2019 Talks

Genetic networks underpinning behavioural changes

Christopher Cunningham

Swansea University, UK

The ability to express multiple behaviours in response to environment change is a fundamental characteristic of animals, especially social animals. The genetic networks that underpin the stable expression of numerous behaviours across multiple animal lineages have been identified over the last twenty years. However, there remains a gap in our understanding about the genetic pathways/networks that are responsible for changing behaviour. Although identifying the genetic networks that underpin behavioural changes has been attempted with a limited number of animals and phenotypes with a few common genes identified across animals. I further addressed this gap by identifying the genetic networks that are activated when an individual changes from a non-parental to a parental state using an insect with complex parental behaviour, including the regurgitation of pre- digested food to dependent and begging offspring. This questions is especially interesting within this species because several well-known mechanisms, cytosine methylation and H3K4me3, appear not to underpin gene expression changes responsible for this behavioural change. Better understanding the mechanistic basis of complex social behaviour and its responsiveness to social environment generates a more complete picture of how animals cope with environmental change helping us better understand the role that behavioural flexibility plays in evolution.

Bayesian inference of natural selection and allele age from time series data of allele frequencies (S)

Zhangyi He, Xiaoyang Dai, Mark Beaumont, Feng Yu

MRC Toxicology Unit, University of Cambridge, UK

Thanks to advances in ancient DNA preparation and sequencing techniques, time serial samples are becoming more and more widely available in ancestral populations. Such time series data allow for more precise inference of population genetic parameters and hypothesis testing on the recent action of natural selection. We develop a novel Bayesian procedure for jointly estimating the selection coefficient and the allele age from the allele frequency time series data sampled from a single locus. Our method uses the particle marginal Metropolis-Hastings to infer the selection coefficient and the allele frequency of the underlying population at the first sampling time point of a non-zero observation, and then estimates the allele age by solving the Kolmogorov backward equation. We illustrate through simulations that our method produces unbiased estimates for the selection coefficient and the allele age. We apply our method to two loci associated with coat colour in horses, previously thought to be under positive selection and linked with domestication.

17 PopGroup52 Oxford 2019 Talks

Towards the understanding of adaptation and speciation in the Swiss Alpine whitefish radiation (S)

Rishi De-Kayne, Ole Seehausen, Philine Feulner

Eawag & University of Bern, Switzerland

The recent origin and repeated diversification of fish in post-glacial lakes make them an ideal system in which to study the genomic changes underpinning ecological diversification and speciation. Within the Alpine whitefish (Coregonus spp.) species complex, the independent diversification into six or more sympatric species of whitefish in multiple lakes across Switzerland are thought to have been facilitated by ecological and reproductive strategy differentiation with sympatric species varying in diet, body length, gill-raker number and spawning depth. This fascinating adaptive radiation provides the opportunity to understand how speciation proceeds in the face of gene flow and the genomic signatures of this process. We have produced two genetic resources to assist investigations into adaptation and speciation in this radiation, a dense linkage map and the first denovo whitefish genome assembly. The linkage map was constructed using one full- sib family including ~150 offspring and contains over 5000 SNPs and suggests that whitefish have multiple lineage specific patterns of chromosome structure. Our whitefish genome was produced from high coverage (~60x) PacBio data, which assembled into a ~3Gb reference, recovering most (92%) known conserved genes. The assembly includes ~60% transposable elements, with class II DNA transposons constituting almost a quarter of the assembly. We now aim to resolve the phylogenetic relationships and determine the prevalence of gene flow between members of the Alpine whitefish radiation using whole- genome sequence data from two individuals of each of the 30 radiating members of the species complex.

18 PopGroup52 Oxford 2019 Talks

Genome-wide association study of carriage versus invasive disease in Neisseria meningitidis

Sarah Earle, Holly Bratcher, Odile Harrison, Christoph Tang, Martin Maiden, Daniel Wilson

University of Oxford, UK

The human pathogen Neisseria meningitidis, the meningococcus, is a major cause of bacterial septicaemia and meningitis worldwide. However, the meningococcus is principally a commensal, and the bacterial genetic factors promoting the transition of asymptomatic carriage to invasive disease remain to be fully elucidated. Host factors such as carriage state, complement deficiencies, social behaviour, and geographic location are associated with increased disease risk. Colonisation with hyperinvasive meningococci is also a major risk factor for invasive disease, but the bacterial genetic mechanisms underlying invasiveness are not well understood. We investigated the bacterial genetic basis of invasive disease at the population level by a genome-wide association study (GWAS) of 261 N. meningitidis isolates. We identified novel associations with invasive disease in fba, the gene upstream to the virulence factor and vaccine component factor H binding protein (fHbp) gene, and hypothesised a mechanism by which the fba variants could affect expression of fHbp. We also recapitulated associations within capsule biosynthesis genes and the Meningococcal Disease Associated island gene tspB. We validated the association in the fba region in a replication set of 1,288 ST-41/44 complex isolates, within which the fba association was a common variant. This has important implications for understanding virulence in N. meningitidis, and why only some strains cause disease. Association studies of this kind have the potential to provide insight into identifying virulent strains, and could further provide candidate targets to assist approaches in treating and preventing meningococcal disease.

19 PopGroup52 Oxford 2019 Talks

A two locus composite-likelihood approach for co-estimating crossing-over and gene conversion rates (S)

Sam Ebdon, Derek Setter, Ben Jackson, Konrad Lohse

IEB, University of Edinburgh, UK

The transfer of genetic variation between homologous chromosomes during meiosis occurs via crossovers and gene conversion. Estimates of recombi-nation rates typically include crossovers, without mention or consideration of gene conversion. As both processes facilitate the exchange of genetic variation, distinguishing crossover and gene conversion rates is essential for a complete understanding of linkage. Existing methods in population genetics use phased SNP pairs to estimate recombination rates (Chan et al., 2012) without estimating gene conversion rates. Population genetics method can estimate gene conversion rates (Gay et al., 2007), albeit with consider- able computational time or only when applied to small samples. Alternatively, recombination rates can be estimated from first principles using two locus theory. Strobeck & Morgan (1978) derived expectations of identity by descent states for for a diploid sample at a fixed distance. These expectations were used by Haubold et al. (2010) in a composite likelihood method. I extended this method to include an effective recombination rate, as a function of both crossing over and gene conversion, originally derived by Frisse et al., (2001). I applied this method to simulations, including gene conversion, and to a well described model species Mus castaneus, to demonstrates the ability of this method to recover expected patterns of recombination rates. I calculated a genome wide estimate of ρx at 0.0014 and ρg at 0.003, showing approximately a 2:1 gene conversion to crossover rate, and demonstrating gene conversion is a major contributor to patterns of linkage disequilibrium we observe in the genome.

20 PopGroup52 Oxford 2019 Talks

An artificial neural networks approach to date ancient skeletons from their DNA

Umberto Esposito, Ghadeer Alshehab , Grace Holland, Eran Elhaik

The University of Sheffield, UK

During the last years, we have witnessed a conspicuous uptick in the sequencing of ancient DNA data, with the proliferation of studies attempting to shed light on human origins and migration routes. Dating the ancient DNA is, therefore, of key importance. However, although radiocarbon dating is a well-established practice, its outcome can be altered by numerous factors, leading to inaccuracies with carbon dates older than 25,000 BP considered highly unreliable. Here, we introduce an artificial neural networks (ANN) approach to date ancient skeletons using genetic information. For that, we curated a dataset of 297 ancient Eurasians with reliable radiocarbon dates ranging from 15,000 BP to 1,500 BP to identify admixture components where allele frequencies vary over time (not geography). We then curated a larger dataset with ~800 ancient Eurasians and split it into two. The ANN algorithm was trained on the training dataset with the admixture components and the known radiocarbon age before it was applied to the test dataset. We show that our predicted dates are perfectly correlated with radiocarbon dates (r^2=0.9). In 90% of the skeletons the predicted age was within 500 years of the average range of radiocarbon dating and in 96.5% it was within 1000 years. Applying our method to skeletons of unknown periods sheds new light on their origins. Our results demonstrate the application of neural networks combined with time-dependent allelic frequency for biodating and contribute to the understanding of the complexity of migration and admixture events.

Quantifying pleiotropy and genotype-by-environment interactions

Tom Ellis, Jon Ågren

Gregor Mendel Institute of Molecular Plant Biology, Vienna

The effect of an allele depends on the context in which it is measured. In particular, an allele might have different effects on seprate traits, or affect the same trait differently in separate environments. Most work in this area has attempted to classify effects based on whether an allele has an effect in the same direction in each context ('positive pleiotropy'), in opposite directions ('antagonistic pleiotropy'), or significant effects on one trait only ('conditional neutrality'). However, this ignores useful quantitative information about the mechanisms of selection, and introduces a well-known statistical bias towards conditional neutrality. I describe a simple method to quantify pleiotropic mechanisms, and apply it to reciprocal transplant experiments investigating the genetic basis of local adaptation. Results indicate that positive pleiotropy is much more common than previously realised, partly because it has been largely ignored.

21 PopGroup52 Oxford 2019 Talks

Genome dynamics after recurrent allopolyploidization and their ecological implications in Dactylorhiza (Orchidaceae) (S)

Mimmi Eriksson, Marie Brandrud, Thomas Wolfe, Ovidiu Paun

University of Vienna, Vienna, Austria

Whole genome duplication (WGD) and hybridization are among the most powerful forces in the evolution of angiosperms. The merging of two divergent genomes in one cell provides a lot of challenges, such as regulatory incompatibilities, reshaping the epigenetic landscape and transposable element expansion. However, there are also positive outcomes such as more allelic diversity, gene redundancy including genetic buffer and possibly larger phenotypic variation and hybrid vigour. For a neopolyploids to succeed in establishing it must either out-competing the parental species or expanding to a new ecological niche. To further understand the underlying forces behind neopolyploid establishment in natural populations the orchid genus Dactylorhiza offers an interesting study system. Allopolyploidization has occurred across the range of the two divergent diploid species D. incarnata and D. fuchsii at multiple times resulting in the origin of sibling allopolyploids adapted to different niches. These sibling alloployploids provides natural replicates at different time points making it possible to investigate how recent allopolyploidy can drive genome evolution, adaptation and provide functional variation leading to colonization of a wide range of niches. Our group are combining RNA-seq data from common garden and reciprocal transplantation experiments, with RAD-seq and low coverage WG data to investigate differential expression, demography and transposable element dynamics between sibling allopolyploids and their parental species.

22 PopGroup52 Oxford 2019 Talks

The effective population size is correlated to census population size in mammalian mitochondrial DNA.

Jennifer James, Adam Eyre-Walker

University of Sussex, UK

The factors that determine the level of genetic variation of a species remain one of molecular evolution’s enduring mysteries. The level of neutral genetic diversity depends upon the mutation rate and effective population size, and the latter might reasonably be expected to be correlated to census population size. However, genetic diversity is generally uncorrelated to any measure of census population size. To help understand what factors determine the level of genetic diversity, we compiled mitochondrial DNA sequence diversity data from >600 mammalian species. Controlling for phylogeny by using paired comparisons we find that synonymous nucleotide diversity is simultaneously positively correlated to range size, and negatively correlated to body size and latitude, with the strongest effect being that of range size. Failing to control for phylogeny yields very different results. To investigate whether the correlation between diversity and range size is due to variation in effective population size, we also analysed a measure of the efficiency of selection – synonymous nucleotide diversity divided by the sum of the synonymous and non-synonymous nucleotide diversities. We find that this measure is negatively correlated to range size. It therefore seems that the effective population size does increase with census population size. However, the relationship is weak both in terms of its slope and the variance explained.

The landscape of coadaptation in Vibrio parahaemolyticus

Daniel Falush1, Chao Yang, Yujun Cui2

1University of Bath, UK 2Institute of Microbiology and Epidemiology, Beijing

We exploit the nearly panmictic population structure of the gastric pathogen Vibrio parahaemolyticus to perform a genome-wide screen for coadapted genetic elements. The great majority of interactions that we detect are between accessory genes, many involved in carbohydrate transport and metabolism. The most complex co-adaptations we identify include hundreds of core genome SNPs and accessory genome elements and involve genes encoding lateral flagella and cell wall biogenesis, implying that several strategies have evolved for colonizing surfaces. Our results provide evidence that, as in human relationships, coadaptation involves progressively increasing levels of commitment, with the most involved interactions becoming irreversible and presaging speciation.

23 PopGroup52 Oxford 2019 Talks

Are supergenes required in intra-specific variation of social organisation?

Emeline Favreau, Claude Lebas, Anurag Priyam, Rodrigo Pracana, Eckart Stolle, Max Reuter, Serge Aron, Yannick Wurm (S)

Queen Mary University of London, UK

Intra-specific phenotypic variation provides an excellent opportunity for understanding the relative strengths of environmental and genetic determinants, as well as characterising the Mendelian elements that may be involved. Here we focus on the intra-specific variation of social organisation in ants, of which the exact evolutionary mechanisms remain largely unknown. Some ant species display a dimorphism of a fundamental social trait, the number of reproductive queens in a colony. Their populations include both types of colonies, with one or multiple queens. Two of these species have independently evolved a large region of the genome associated with colony social form. In each species, a pair of non-recombining "social chromosomes" (supergene) determines whether a colony contains a single queen or multiple queens. We test whether such architecture is also involved in a third, unrelated lineage where social dimorphism convergently evolved. We find that 4% of all biallelic SNPs in the genome are associated with social type. We mapped those loci to a de novo assembly and assessed their linkage to determine if they indeed form a supergene. These results will inform us about the molecular constraints underlying convergent evolution of major social phenotypes.

Detecting selection on structural variants from the linked frequency spectrum

Luca Ferretti, Carla Giner-Delgado, Alex Klassmann, Thomas Wiehe, Sebastian Ramos- Onsins, Mario Cáceres, Guillaume Achaz

The Pirbright Institute and Big Data Institute, UK

Both SNPs and structural variants (SVs) are important sources of genomic and phenotypic variability in populations, often with strong phenotypic effects. For this reason, many polymorphic SVs in natural population could be under selective pressure. However, detecting ongoing selection on SVs is a challenging task due to their impact on local patterns of variation and recombination. Here we present a new set of methods to detect violations of neutrality in several simple SVs, including inversions, deletions and insertions, as well as introgressions from close species. These methods are based on the linked frequency spectrum and can detect both positive and balancing selection on SVs. We present some applications of these methods to SNPs and SVs in human populations.

24 PopGroup52 Oxford 2019 Talks

Global genome diversity, hybridisation & aneuploidy in the Leishmania donovani complex

Susanne U. Franssen1, Caroline Durrant1, Olivia Stark2, Bettina Moser2, Mandy Sanders1, Hideo Imamura3, Tim Downing4, Jean-Claude Dujardin3, Thierry Wirth5, Matthew Berriman1, Gabriele Schönian2, James A. Cotton1

1Wellcome Sanger Institute, Hinxton, UK 2Charité Universitätsmedizin, Berlin, Germany 3Institute of Tropical Medicine, Antwerp, Belgium 4Dublin City University, Dublin, Ireland 5Muséum National d'Histoire Naturelle, Paris, France

The genus Leishmania are unicellular protozoan parasites that cause leishmaniasis, a neglected tropical disease that is prevalent in developing countries. Parasites are transmitted by sandflies and replicate intracelluarly in macrophages of human or other mammalian hosts. Two interesting features shape the Leishmania genome: First, even though they mainly reproduce clonally, hybrids within and between species have been observed. Second, Leishmania species show mosaic aneuploidy, where individual chromosomes frequently experience copy number variation even within a strain. We investigate the genome-wide diversity of 151 globally distributed natural isolates from the L. donovani complex (comprising L. donovani and L. infantum) predominantly causing the fatal disease visceral leishmaniasis. Phylogenetic reconstruction separates L. donovani samples into four diverse subgroups roughly coinciding with geographical origin. In contrast, little diversity is observed for globally distributed L. infantum samples. We find evidence for past hybridisation in some samples with high heterozygosity, while others show elevated heterozygosity without putative parental samples to support hybridisation. In contrast, 60% of the samples are almost entirely homozygous. Aneuploidy patterns across samples are inconsistent with the genome-wide phylogeny in accordance with a high turnover rate of aneuploidy. Using individual clades as replicates, we observe chromosome-specific copy number and somy variability. Chromosomes with elevated variability typically show reduced heterozygosity, providing the first experimental evidence for the reduction of heterozygosity due to aneuploid turnover. The results suggest profound impacts of interplay between facultative sexuality and highly variable aneuploidy in Leishmania that make the population genetics of this human pathogen highly unusual.

25 PopGroup52 Oxford 2019 Talks

Quantifying natural selection from genomic data using deep learning

Luis Torada, Ulas Isildak, Lucrezia Lorenzon, Alice Beddis, Sara Mathieson, Matteo Fumagalli

Imperial College London, UK

The identification of selection signatures in the genome has the potential to unveil the genetic mechanisms underpinning adaptation. Commonly used strategies to detect selection signals from genomic data rely on compressing the information into summary statistics. Furthermore, few methods are able to quantify the strength of selection. Here we explored the use of deep learning to make full use of population genomic data and developed a supervised machine learning algorithm for an accurate estimation of selection coefficients. We represent the information of genomes from multiple individuals as abstract images, depicting the genomic variation at each position for each sample. To quantify signatures of positive selection, we implemented a convolutional neural network to detect patterns of selection as a local reduction of genetic variation. We trained the network using extensive simulations and estimated the selection coefficient using multi-classification techniques. We show how our method, implemented in the pipeline ImaGene, can be affected by data manipulation, learning strategies, and demographic model specification. Nevertheless, our method is able to generate reliable probability distributions for selection coefficients. We finally demonstrate how this approach can be extended to understand other selective events, including balancing selection. While the use of deep learning in evolutionary genomics is in its infancy, here we demonstrated its potential to detect informative patterns from large-scale genomic data. The joint inference of the evolutionary history of mutations and their functional impact will provide us with novel insights into the molecular mechanisms associated with adaptation.

26 PopGroup52 Oxford 2019 Talks

Mutation load dynamics during environmentally-driven range shifts

Kimberly J. Gilbert, Stephan Peischl, Laurent Excoffier

University of Bern, Switzerland

The fitness of spatially expanding species has been shown to decrease over time and space, but specialist species tracking their changing environment and shifting their range accordingly have been little studied. We use individual-based simulations and analytical modeling to compare the impact of range expansions and range shifts on genetic diversity and fitness loss, as well as the ability to recover fitness after either a shift or expansion. We find that the speed of a shift has a strong impact on fitness evolution. Fastest shifts show the strongest fitness loss per generation, but intermediate shift speeds lead to the strongest fitness loss per geographic distance. Range shifting species lose fitness more slowly through time than expanding species, however, their fitness measured at equal geographic distances from the source of expansion can be considerably lower. These counter-intuitive results arise from the combination of time over which selection acts and mutations enter the system. Range shifts also exhibit reduced fitness recovery after a geographic shift and may result in extinction, whereas range expansions can persist from the core of the species range. The complexity of range expansions and range shifts highlights the potential for severe consequences of environmental change on species survival.

Unveiling the genetic history of the Maniq, a pristine hunter-gatherer population of Southeast Asia (S)

Tobias Göllner, Maximilian Larena, Helmut Lukas and Helmut Schaschl

University of Vienna, Vienna, Austria

Southeast Asia is inhabited by various indigenous human populations which show huge morphological, anthropological, and linguistic diversity. However, we have only limited knowledge about the evolutionary history of these populations. Part of this indigenous people in Southeast Asia are the Maniq. They are one of the few remaining primary hunter-gatherers of the world. The Maniq are the northernmost Semang and, unlike the other Semang, live in the rainforest of Thailand (Provinces: Trang, Phattalung and Satun). Recently, long-term participant researcher Helmut Lukas received ten saliva samples from them. In this project, we want to uncover the genetic ancestry of the Maniq, understand their phylogenetic position among human populations and investigate the genetic adaptation to the hunter-gatherer lifestyle. The samples were analysed using the Infinium Omni2.5-8 Exome v1.3 BeadChip which gives us information on approximately 2,6 million markers from autosomes, sex chromosomes and mitochondrial DNA. We present autosomal data on the Maniq in reference to selected populations from 1000 Genomes, SGDP, and shared data from publications. Initial analyses focus on Principal Component Analysis (PCAs), phylogenetic trees and FST measurements. Furthermore, we want to look at runs of homozygosity, time to most recent common ancestor and admixture, including Neandertals and Denisovan. Overall, this project will contribute to the ongoing discussion on the peopling of Southeast Asia.

27 PopGroup52 Oxford 2019 Talks

Identification of genes and non-coding conserved regions underpinning bird beak shape evolution on a macroevolutionary scale

Leeban Yusuf, Matthew Heatley, Joseph Palmer, Henry J. Barton, Christopher R. Cooney, Toni I. Gossmann

University of Sheffield, UK

Little is known about the genetic underpinnings of trait evolution on a macroevolutionary scale. A pertinent example of this is beak shape, a key evolutionary innovation linked to macroevolutionary patterns of speciation. Progress has been made identifying genes implicated in the evolution of beak shape on a microevolutionary scale, but the way these genes may be linked to beak shape evolution over macroevolutionary time has not yet been studied. Here, we incorporate geometric-morphometric analyses of beak shape evolution into evolutionary analyses of 10,322 protein-coding genes and XYZ regulatory regions, including known candidate genes, spanning 58 bird species. We detected 36 genes with a signature of positive selection, two (BGLAP and SOX5) of which are strong candidates for lineage-specific beak shape evolution. Furthermore, we found 1434 genes with significant variation in their protein evolutionary rates to explain the large variation in beak morphology observed across the bird phylogeny. Among these, molecular rates of 16 genes show a strong correlation with beak shape change across distantly-related avian taxa, suggesting potential repeated evolution of these genes associated with beak shape. Together, our findings suggest large-effect genes are important for the evolution of key innovations over broad evolutionary time scales. In particular, we find strong evidence for lineage-specific evolution of protein-coding genes possibly linked to beak shape and a range of other adaptations. Finally, we propose that putative large-effect genes undergoing repeated evolution may not play a substantial role in the evolution of complex traits, or may be too difficult to detect using currently available methods and datasets.

28 PopGroup52 Oxford 2019 Talks

Signatures of selective sweeps with arbitrary dominance and self–fertilisation

Matthew Hartfield1,2,3 and Thomas Bataillon2

1Department of Ecology and Evolutionary Biology, Toronto, Canada 2Bioinformatics Research Centre (BiRC), Aarhus, Denmark 3Institute of Evolutionary Biology, Edinburgh, UK

Searching for genetic signatures of adaptation, also called 'selective sweeps', forms a major part of evolutionary genomics research. Many mathematical models have been developed to determine sweep signatures from sequence data. Yet existing models generally assumed that populations are fully outcrossing, and mutations have additive dominance. Self-fertilisation, where offspring inherit two gametes from the same parent, alters selective sweeps in several ways; the expressed fitness dominance of adaptive mutations is also affected by self-fertilisation rates. We build coalescent models of how adaptive mutations affect linked neutral variation, assuming arbitrary dominance and self- fertilisation. Self-fertilisation produces very distinct sweep signatures, as adaptive variants fix in a shorter period of time, and lower effective recombination rates create distinct haplotypes following the sweep. Dominance only strongly affects selective sweeps in outcrossing organisms; adaptation from standing variation further weakens the influence that dominance exerts on signatures of adaptation. We demonstrate the applicability of these models by applying them to sweep signatures in humans and the selfing plant Arabidopsis thaliana, to determine the underlying adaptation mechanisms in both cases. These models offer a more realistic basis for quantifying the type and strength of adaptive mutation in a broad range of organisms.

29 PopGroup52 Oxford 2019 Talks

An MCMC-based method for estimating selection coefficients from time series DNA data across linked loci

Zhangyi He, Xiaoyang Dai, Mark Beaumont, Feng Yu

MRC Toxicology Unit, University of Cambridge, UK

The rapid improvement of DNA sequencing technology has made it possible to monitor genomes in great detail over time, which presents an opportunity for investigating natural selection based on time series DNA data while accounting for genetic recombination. Such time series data allow for more precise estimates for population genetic quantities and hypothesis testing on the recent action of natural selection. Here we develop a novel Bayesian statistical framework for the inference of natural selection by capitalising on the temporal aspect of DNA data across linked loci, regardless of whether the genotype data have been phased or not. Our Bayesian inference procedure relies on the hidden Markov model incorporating the two-locus Wright-Fisher diffusion with selection, which enables us to take genetic recombination into account explicitly, and the posterior probability distribution for the selection coefficients is achieved through the particle marginal Metropolis-Hastings, which allows for co-estimation of the haplotype frequency trajectories of the underlying population. The performance of our method is evaluated through extensive simulations, and we show that our estimates for the selection coefficients are unbiased. Moreover, under certain circumstances, we find that our approach can deliver precise estimates for the selection coefficients whereas existing single locus methods fail, especially when the loci are tightly linked. The utility of our approach is illustrated with an application to the linked loci encoding coat colour in horses, where we also show that our method is able to handle missing values in DNA data.

30 PopGroup52 Oxford 2019 Talks

From boom to bust – bacterial adaptation to prolonged survival following resource exhaustion

Sarit Avrani, Evgeni Bolotin, Sophia Katz and Ruth Hershberg

Technion - Israel institute of Technology, Israel

Many bacteria, including the model bacterium Escherichia coli can survive for years within spent media, following resource exhaustion. We carried out evolutionary experiments, followed by full genome sequencing of hundreds of evolved clones to study the dynamics by which E. coli adapts during the first four months of survival under resource exhaustion. Our results reveal that bacteria evolving under resource exhaustion are subject to intense selection, manifesting in rapid mutation accumulation, enrichment in functional mutation categories and extremely convergent adaptation. Our results further demonstrate that such adaptation is not limited by mutational input. Indeed, mutational input appears to be high enough to enable bacteria to rapidly adapt, in a highly convergent manner and with great temporal precision through fluctuations in allele frequencies. Finally, we demonstrate that due to antagonistic pleiotropy and mutation accumulation, survival under resource exhaustion can severely reduce a bacterium’s ability to grow exponentially, once resources are again available. We study how this affects the ability of bacterial populations to re-adapt to growth once resources again become available. Combined, our results shed light on bacterial adaptation to long-periods of resource exhaustion and on the consequences such adaptation has on the genetic makeup of individual bacteria and on patterns of genetic variation within bacterial populations.

Alternative adaptation to a long term DNA virus infection in Drosophila innubila

Tom Hill, Robert L. Unckless

University of Kansas, USA.

Viruses make up a considerable proportion of the pathogens infecting animals. They can spread rapidly between hosts and sicken or even kill their host to promote their own proliferation. Due to this strong selective pressure, antiviral immune genes are some of the fastest evolving genes across metazoans, as highlighted in mammals and insects. While Drosophila melanogaster are frequently exposed to pathogenic RNA viruses, little is known about D. melanogaster’s ecology, or if they are representative of other Drosophila species in terms of pathogen pressure. Here, we sequence and assemble the genome of a highly diverged, mushroom-feeding Drosophila species, Drosophila innubila, a species frequently exposed to a highly pathogenic DNA virus. We investigate the evolution of Drosophila innubila, focusing on the immune system and find little evidence for adaptive evolution of the antiviral RNAi genes, though we do find several other immune pathways are evolving adaptively, suggesting alternate means of viral tolerance including some antiviral non-RNAi pathways characterized in D. melanogaster. This suggests that evolution of tolerance to DNA viruses may differ from that of RNA viruses and is conserved across Drosophila. Additionally, we find unexpected population structure between locations for both the host and virus, suggesting flies are travelling much further than previously thought between populations.

31 PopGroup52 Oxford 2019 Talks

Call 999 and ask for population genetics; rapid turnaround analysis of viral sequence data

Casper K. Lumby, Lei Zhao, Christopher J.R. Illingworth

University of Cambridge, UK

We describe the rapid analysis of viral sequence data from a chronically infected influenza patient with an immunodeficiency. We show how drug therapy shaped the pattern of influenza infection over a long period of treatment, and evaluate the use of population genetic techniques for gaining insight into the composition of the viral population and into the overall patten of viral evolution. We discuss the potential for population genetics to be applied in a clinical context as a diagnostic procedure for patients and as a means for improving the treatment of long-term viral infections.

Haploid selection in a predominantly diploid animal

Ghazal Alavioon, Roy Francis, Cosima Hotzy, Claire Armstrong, Khriezhanuo Nakhro, Douglas Scofield, Julia Wyszkovska, Alexei, Maklakov, Lewis Spurgin, Simone Immler

University of East Anglia, UK

A consequence of sexual reproduction in eukaryotes is the evolution of a biphasic life cycle with alternating diploid and haploid gametic phases. While our focus in evolutionary biology is on selection during the diploid phase, we know relatively little about selection occurring during the haploid gametic stage. This is particularly true in predominantly diploid animals with a short haploid gametic phase. We tested the idea of selection during the haploid gametic phase in the zebrafish. We combined a large scale selection experiment over three generations with fitness assays and G&T sequencing to assess the importance of haploid selection. In the experiment, we selected on sperm phenotypes by manipulating the time between sperm activation and fertilisation with two treatments: short- activation time (SAT) and long activation time (LAT) distinguish between short- and long- lived sperm within the ejaculate of a male. Offspring sired by LAT sperm exhibited overall higher fitness compared to their siblings sired by SAT sperm. Comparing gene expression between the treatments by RNA sequencing of brain and testes in F1 and F3 offspring showed differentially expressed genes in metabolic and developmental pathways potentially causing the observed fitness differences. Furthermore, we sequenced SAT and LAT sperm pools from the same ejaculate from three different males and identified allelic differences between the pools. In addition, we sequenced F1 offspring and detected surprisingly strong signals of selection supporting the idea that haploid selection plays an important role even in diploid animals.

32 PopGroup52 Oxford 2019 Talks

Genomic features of asexual animals (S)

Kamil S. Jaron, Jens Bast, Marc Robinson-Rechavi, Tanja Schwander

Department of Ecology and Evolution, University of Lausanne, Lausanne, Switzerland Swiss Institute of Bioinformatics, Lausanne, Switzerland

Evolution under asexuality is predicted to impact genomes in numerous ways, but empirical evidence remains equivocal. While descriptions of peculiar genomic features of single asexual animal lineages exist, it is unclear how general these observations are. Therefore, we reviewed and reanalyzed published data of 24 asexual species from five different phyla, including chordates, rotifers, arthropods, nematodes and tardigrades. We recomputed features hypothesized to be the most affected by asexuality, using consistent methods. We uncovered a number of unusual features in the genomes of asexual animals. However, none of these were systematically replicated across the majority of analyzed species, suggesting there is no universal genomic feature specific to asexual species. We found that a hybrid origin of asexuality was the most important factor affecting heterozygosity, with potential effects of asexuality being masked by effects of hybrid ancestry. Asexuals that are not of hybrid origin appeared to be largely homozygous, independently of the cellular mechanism underlying asexuality.

A Young Social Chromosome

Max John, Benjamin Braim, Robert Hammond

University of Leicester, UK

Supergene architectures are increasingly understood to control inheritance of a variety of complex traits in diverse eukaryotic genomes. Questions exist about the selection regimes which favour supergene formation, about the genes and traits which are susceptible to being captured by expanding non-recombining regions, and about the ubiquity of supergenes. The study of genetic control of social organisation in social insects is privileged to be able to address all of these questions simultaneously. Recent work has demonstrated the convergent evolution of “social chromosomes” controlling a within- population social polymorphism, in two divergent ant species. We present an analysis of the genomic architecture of a distinct, between-population social polymorphism in a third ant, Leptothorax acervorum. Using a combination of whole genome sequencing, de novo assembly and RADseq-based population genomics, we have identified >200 SNPs strongly associated with fixed variation in social organisation between populations. A preliminary investigation of long-range genomic structure shows that many of these markers cluster in a large region of a single chromosome. This region is not syntenic to either of the previously identified social chromosomes and, although divergence is elevated relative to the rest of the genome, in absolute terms it is low. Together these results suggest an independently evolved supergene in L. acervorum, much younger than those discovered in other ant species. Detailed analyses based on a more accurate assessment of the long-range genomic structure of this region is forthcoming. This will allow an important and insightful comparison of three independent supergene structures underpinning inheritance of colony-level adaptations.

33 PopGroup52 Oxford 2019 Talks

Distribution of fitness effects of new mutations in Chlamydomonas

Peter Keightley, Katharina Boendel, Susanne Kraemer, Deirdre McClean, Toby Samuels, Jarrod Hadfield, Rob Ness, Nick Colegrave

University of Edinburgh, UK

We conducted a mutation accumulation (MA) experiment for c.1,000 generations in the single-celled alga Chlamydomonas reinhardtii, and measured growth rate and competitive fitness of evolved lines and their unmutated ancestors. MA lines generally declined in fitness. We sequenced the genomes of the MA lines to identify the molecular basis of the mutations. Fitness and the total numbers of accumulated mutations and coding mutations in the genome were significantly negatively correlated. We fitted a model of the relationship between mutation number and fitness, and found that models containing one or two mutational effect categories (one neutral and one deleterious category) fitted the data best in most cases. We crossed the MA lines to their respective unmutated ancestor to produce backcross lines that have different combinations of the accumulated mutations. This should in principle allow the individual effects of mutations and the distribution of effects to be estimated. We are developing a MCMC approach for this purpose, and the results of applying it to the backcross data will be reported.

Tree sequence recording opens new horizons for forward-time simulations

Jerome Kelleher

University of Oxford, UK

Forwards in time simulations are an essential tool in applied population genetics, allowing us to create realistic simulations of the organisms and processes we study. Although forward simulations are extremely flexible, there has classically been a large cost in terms of CPU time for such realism. In practical terms, models with realistic population sizes and genome sizes have been impossible to run. We discuss the recently developed "tree sequence recording" method which removes this practical barrier and allows us to simulate vastly larger and more complex models. In this method we record the full genealogy of all individuals that every lived in the simulation, periodically discarding ancestry that is not relevant to the extant population. Because we have the complete genealogy, the forward simulation no longer needs to generate and track neutral mutations, which, by definition, do not affect the trees. This removes a large computational burden from the simulator, as well as returning a great deal more information to the user. The ancestry information returned is in the same format as used by the msprime simulator, and the same code can be used to analyse tree sequences generated by both backwards and forwards time simulators. We show some practical examples from the SLiM simulator (which implements this method), in which some models run more than a million times faster due to tree sequence recording.

34 PopGroup52 Oxford 2019 Talks

Detection of strong decline in populations by genomic approaches (S)

Elise Kerdoncuff1,2, Amaury Lambert2,3, Guillaume Achaz1,2

1Institut de Systématique, Evolution, Biodiversité (ISYEB), MNHN, Paris, France 2Centre Interdisciplinaire de Recherche en Biologie (CIRB), Collège de France, Paris, France 3Laboratoire de Probabilités et Modèles Aléatoires (LPMA), UPMC, Paris, France

Only 5% of described species have a conservation status. Methods used to assess conservation status cannot be generalized to all species. We thus developed a model that predicts the DNA sequences of a sample from a population evolving under different demographic scenarios. Using this model, we studied features of DNA sequences under a constant size or a strong population decline. We developed a new method to study demography based on the length of compatible blocks along the genome, e.g. blocks of nucleotides within which we cannot detect recombination events. Lengths of compatible blocks depend on the frequency of recombination events which is influenced by the ancestral history of the population. This method allows us to infer very recent decline of a population from DNA sequences that could be a new tool to assess conservation status in a wide kinds of life organisms.

Robust estimation of recent effective population size from number of independent origins in soft sweeps

Bhavin S. Khatri and Austin Burt

Imperial College, UK

Estimating recent effective population size is of great importance in characterising and predicting the evolution of natural populations. Methods based on nucleotide diversity may underestimate current day effective population sizes due to historical bottlenecks, whilst methods that reconstruct demographic history typically only detect long-term variations. However, soft selective sweeps, which leave a fingerprint of mutational history by recurrent mutations on independent haplotype backgrounds, holds promise of an estimate more representative of recent population history. Here we present a simple and robust method of estimation based only on knowledge of the number of independent recurrent origins and the current frequency of the beneficial allele in a population sample, independent of the strength of selection and age of the mutation. Using a forward time theoretical framework, we show the mean number of origins is a function of θ=2Nµ and current allele frequency, through a simple equation, and the distribution is approximately Poisson. This estimate is robust to whether mutants pre-existed before selection arose, and is equally accurate for diploid populations with incomplete dominance. For fast (e.g., seasonal) demographic changes compared to time scale for fixation of the mutant allele, and for moderate peak- to-trough ratios, we show our constant population size estimate can be used to bound the maximum and minimum population size. Applied to the Vgsc gene of Anopheles gambiae, we estimate an effective population size of roughly 6x10^7, and including seasonal demographic oscillations, a minimum effective population size greater than 6x10^6 and a maximum less than 3x10^9.

35 PopGroup52 Oxford 2019 Talks

Slow Y-degeneration and early rise of dosage compensation on Silene sex chromosomes.

Marc Krasovec1, Kate Ridout1, Yusuke Kazama2, Kotaro Ishii2, Michael Chester1, and Dmitry A. Filatov1

1Department of Plant Sciences, University of Oxford, Oxford OX1 3RB, United Kingdom 2RIKEN Nishina Center for Accelerator-Based Sciences, 2-1 Hirosawa, Wako, Saitama, 351-0198, Japan

Y degeneration, where genes on the Y chromosome become non-functional over time, may be much slower in plants due to strong purifying selection against deleterious mutations in the haploid gametophyte. Testing for differences in Y degeneration between the kingdoms has been hindered by the absence of accurate age estimates for plant sex chromosomes. We used genome resequencing in parents and progeny to estimate the spontaneous mutation rate and the age of the sex chromosomes in white campion (Silene latifolia), which yielded a rate of 7.31E-09 (95%CI: 5.20E-09 - 8.00E-09) mutations per site per generation. Given this mutation rate, the old and young strata of S. latifolia sex chromosomes are 11.00 and 6.32 million years old. The analysis of Y-degeneration revealed that at least 47% of S. latifolia Y-linked genes are already dysfunctional and the rate of Y degeneration in S. latifolia is nearly two-fold slower than in animals. Y- degeneration creates imbalance of gene product between the sexes, which is compensated by dosage compensation (DC) system. Previous studies reported the presence of a partial DC in S. latifolia, but it is not known whether that DC has to evolve gene-by-gene or some form of chromosome-wide dosage compensation is already present in this species. To shed light on this question we used artificially created Y- deletion mutants to manipulate gene dose and analyse gene expression in S. latifolia. We demonstrate that deletions of Y-linked genes result in immediate up-regulation of X-linked gametologs, indicating the presence of chromosome-wide dosage compensation system that arose surprisingly early in sex chromosome evolution in S. latifolia.

36 PopGroup52 Oxford 2019 Talks

Population genomics of swallowtail butterflies: Iphiclides podalirius and I. feisthamelii

Dominik R. Laetsch, Roger Villa, Alex Hayward, Konrad Lohse

University of Edinburgh, UK

The scarce swallowtail butterfly Iphiclides podalirius [Linnaeus, 1758] and I. feisthamelii are parapatric siblings in Europe. The two species display subtle differences in genital morphology, wing patterns, and UV reflection in males, yet still hybridise in a narrow contact zone close to the Pyrenees in Southern France. Here, we present the genome of I. podalirius assembled from short (Illumina) and long (PacBio and 10x Chromium) read data to an scaffold N50 of 4 Mb. We re-sequenced five individuals of both species – sampled both close and far from the contact zone – and developed a novel bioinformatic toolkit to calculate block-wise site-frequency-spectra (bSFS) from multisample variant data. We analyse bSFS data in a coalescent framework to infer the demographic history of the species pair and describe the degree of and heterogeneity in differentiation along the genome. Specifically, we i) ask whether divergence is compatible with vicariance into distinct atlantic (I. feisthamelii) and south mediterranean (I. podalirius) refugia during the last glacial maximum, and ii) interrogate the samples in respect to their Wolbachia endosymbiont strains.

Genetic and phenotypic diversity of brown trout (Salmo trutta) populations in Iceland (S)

Lagunas M. G., Pálsson A., Jónsson Z.O. and Snorrason S.S.

Faculty of Life and Environmental Sciences - University of Iceland, Ireland Populations of brown trout in several watersheds of Iceland may have been isolated early in post-glacial history (ca. 10,000 years ago) by the formation of impassable waterfalls or due to other topographic changes to waterways. This extended isolation from populations living downstream of impasses is expected to have caused considerable genetic differentiation by genetic drift. Considering the variable environments isolated trout populations inhabit, the role of local adaptation may also be significant. The aims of the project are: (i) to assess to what extent the current pattern of genetic relatedness of brown trout populations reflects the topology of the waterways and physical barriers to migration, and (ii) to look for signs of local adaptations. We sampled populations from several watersheds around Iceland and employed ddRADseq to identify single nucleotide polymorphism loci that could be used to assess the genetic divergence of populations. Morphometric analyses were also conducted to evaluate phenotypic differences among populations or between contrasting environments, which may indicate local adaptations. The data show that the pattern of genetic relatedness of populations (estimated by Fst) reflects the topology of waterways and confirm the assumption that the upstream populations have been isolated for a long time. A preliminary morphological analysis suggests a distinction between fish from riverine and lacustrine habitats. By combining these data we hope to find genetic markers associating with phenotypic differences. Once the genetic structure of the populations is determined at a finer resolution we could test hypotheses related to the colonization history of the watersheds.

37 PopGroup52 Oxford 2019 Talks

Fine mapping without phenotyping: identification of selection targets in Evolve and Resequence experiments (S)

Anna Maria Langmüller, Christian Schlötterer

Institut für Populationsgenetik, Vetmeduni Vienna; Vienna Graduate School of Population Genetics, Vienna

Evolve and Resequence (E&R) – the combination of Next Generation Sequencing and Experimental Evolution – offers the possibility to study adaptation under standardized laboratory conditions on the genomic level. Computer simulations suggested that E&R studies using sequencing of pooled individuals (Pool-Seq) can identify genes contributing to adaptation. Nevertheless, E&R studies typically suffer from an excess of candidate SNPs caused by linkage disequilibrium with low frequency selected alleles. Even highly replicated experiments carried out for more than 100 generations are frequently not powerful enough to identify the selected gene or even the causative variant. E&R studies, which do not apply truncating selection, but build on the laboratory natural selection paradigm where fitter individuals have more offspring, suffer from the additional challenge that the selected phenotype is not known. Since, the identification of the selection targets is key to understand the genetic basis of adaptation processes, alternative approaches are needed to fine map selected loci. We used computer simulations to explore a broad range of experimental designs to fine map selection targets even without knowledge about the adaptive phenotype. We show that by means of a second round or E&R experiments with moderate replication selection targets can be mapped with a very high resolution: the causative SNP is under the top 10 significant outliers in more than 70% of the cases.

38 PopGroup52 Oxford 2019 Talks

Measuring the fitness cost of insecticide resistance with Evolve and Resequence: A case study with Ace resistance in Drosophila simulans (S)

Sonja Lecic, Rodolphe Poupardin, Viola Nolte, Agnes Jónás, Christian Schlötterer

Institut für Populationsgenetik, Vetmeduni Vienna; Vienna, Austria

Natural populations rapidly acquire resistance to insecticides. Frequently, this resistance comes with a considerable cost, which is evident from the presence of susceptible and resistance alleles in natural populations. Because the cost of insecticide resistance is difficult to measure in the field, we used experimental evolution to quantify the cost of insecticide resistance. After >100 generations in an insecticide free environment a highly resistant D. simulans population lost resistance to the insecticides Malation and Propoxur. Whole genome sequencing of pools of individuals (Pool-Seq) uncovered a pronounced selection signature around the Ace locus, a classic resistance gene. Three amino acid replacements (I116V, G265A and F330Y), which confer insecticide resistance, decreased in frequency from 50-55% to 0-10%. Using time series data, we inferred a selection coefficient of at least 0.07, confirming the high cost of insecticide resistance. Computer simulations confirmed a dominant or codominant model of Ace resistance. Sequencing 34 haplotypes from the ancestral and 24 haplotypes from the evolved flies, we show that all resistance alleles are located on the same haplotype. This haplotype has low levels of sequence variation, confirming strong directional selection for insecticide resistance in natural D. simulans populations. Similar to D. melanogaster, the same haplotype with three resistance alleles is also detected in other populations, suggesting a global spread, rather than independent origin.

The evolution and function of DNA methylation across arthropods

Samuel H. Lewis

University of Cambridge. UK

DNA methylation is an important modification that can affect the regulation and structural integrity of the genome. Previous studies have shown that the level of DNA methylation varies widely between different insect species, and uncovered phenotypic effects of altered methylation levels in some species. However, we still do not know how widespread DNA methylation is across arthropods, how variable the targets of methylation are between species, and what the relationship is between methylation and the activity of these targets. We used whole genome bisulfite sequencing to quantify DNA methylation at base-pair resolution across 14 arthropod species. We find that DNA methylation is ancestral to all arthropods, and has been lost independently in flies and some crustaceans. We also discover that genes are a conserved target of methylation, and other features are methylated in specific lineages. Finally, we find that highly-methylated features are generally expressed at a higher level than lowly-methylated features. These results suggest that DNA methylation evolves rapidly and dynamically, and interacts with transcription to regulate the activity of arthropod genomes.

39 PopGroup52 Oxford 2019 Talks

Conservation of adaptive potential and functional diversity: how far will genomics take us?

Barbara Mable

University of Glasgow, UK

Continuing advances in whole-genome scale approaches integrated with other "omic” technologies promise to revolutionise understanding about the relevance of genetic variation to risks of species declines and extinctions. In the face of the vastly increased accessibility of such approaches, it is important that we advance beyond descriptive genetics to developing a more functional perspective on whether enhancing genetic variation is the most effective strategy for conservation management. Combining historical and recent examples, this talk will focus on the following inter-related issues: 1) the relative impacts of inbreeding and outbreeding on fitness and adaptive potential, particularly in relation to genetic rescue; 2) how “species” should be defined for conservation management; 3) deciding on how much and what type of genetic variation should be preserved; and 4) how we can move from descriptive genetics to actually understanding adaptive processes in the wild. None of the ideas presented are new; the purpose here is to serve as a reminder that many of the issues raised thirty years ago are still relevant but not completely resolved and would benefit from tackling afresh with modern tools, but combining short-term ecological with long-term evolutionary perspectives.

Is genomic architecture, not evolutionary conflict, the main driver of expression patterns in a social supergene? (S)

Carlos Martinez-Ruiz, Richard A. Nichols and Yannick Wurm

Queen Mary University of London, UK

It is becoming increasingly clear that supergenes play a key role in defining and maintaining complex polymorphisms across a wide range of organisms. We still know little, however, about the evolutionary forces underpinning their evolution. Evolutionary conflict has been proposed as a potential driver of supergene formation, based on our understanding of sex chromosome evolution. Here we focus on the fire ant Solenopsis invicta system to test this idea. This ant species has a ‘social chromosome’ S, which contains an inversion polymorphism (SB or Sb) that acts as a supergene. The social chromosome determines whether a colony has one or multiple reproductive queens, two social forms that differ in multiple key ecological traits. Recombination between SB and Sb is severely repressed and, since the Sb homozygotes are lethal recessives, there is no effective recombination within Sb. This genetic architecture is very similar to that of sex chromosomes, which also have variants that are essentially non-recombining (the Y in an XY system and the W in a WZ system). We hypothesise that the evolution of the social chromosome system has been shaped by social antagonism between the two social forms, in the same way that sexual antagonism shapes the evolution of sex chromosome pairs. We test whether expression patterns in the social chromosome match those expected by evolutionary conflict, as happens in sex chromosomes. Additionally, we model the gene flow between the two social forms to determine the potential importance of social antagonism in this system. Our results are consistent with socially antagonistic selection being responsible for the divergence of the social chromosome. These findings will shed general light on the evolution of supergene regions of suppressed recombination.

40 PopGroup52 Oxford 2019 Talks

Hybridization fuels repeated bouts of adaptive radiation in Lake Victoria Region cichlid fishes

Joana Meier, Ole Seehausen

University of Cambridge, UK

Of the four cichlid lineages that colonized Lake Victoria, three did not speciate and one radiated into 500 species. The same lineage which radiated in Lake Victoria also formed adaptive radiations in all other major lakes in the region. These radiations together have been termed “Lake Victoria Region Superflock” (LVRS) encompassing 700 species distributed across four major and many smaller lakes. We have previously shown that the entire LVRS is derived from a hybrid swarm of two divergent cichlid lineages that fueled the adaptive radiations 150,000 years ago. However, the current Lake Victoria radiation is much younger than the LVRS, because the lake dried up and refilled only 15,000 years ago. Despite its youth, the Lake Victoria radiation is highly diverse in phenotype and ecology ranging from specialized algae scrapers of rocky shores to demersal detritivores and large piscivores. We study if these different ecological groups evolved independently in Lake Victoria or if each of them colonized Lake Victoria from one of the older radiations in the LVRS. We find that at least two LVRS species colonized the modern lake and admixed prior to radiating into 500 species. We provide evidence that the different ecological groups have arisen in situ, but through sorting of old variation which was repeatedly reshuffled through hybridization and sorted into the different ecological groups and species.

41 PopGroup52 Oxford 2019 Talks

Population genetic survey of wild Brassica oleracea in the UK and Spain: unravelling the origins of an important crop wild relative (S)

Elizabeth Mittell, Barbara Mable, Christina Cobbold, Umer Ijaz

University of Glasgow, UK

Domestication is fascinating to both scientists and non-scientists, as it not only revolutionized human societies, but it also dramatically altered the evolutionary trajectories of domesticated species. A classic example of domestication, recognised by Darwin, can be found in cabbages (Brassica oleracea). This single species contains a huge amount of morphological diversity in cultivated varieties, such as kale, kohlrabi, broccoli, Brussels sprouts and cauliflower, which is not found in wild populations. Due to the agricultural importance of cabbages, the issue of food security, and the idea that crop wild relatives hold important genetic resources ex situ, understanding the domestication history and genetics of wild populations of cabbages has become economically, as well as scientifically, relevant. To date, the most plausible hypothesis is that these crops were domesticated in the Mediterranean region, before being moved across Europe by people, where escaped plants established populations across the Atlantic region. Although this is supported by historical records and some genetic evidence, a comprehensive genetic analysis of these wild populations has been lacking. Here, sampling was carried out across a latitudinal gradient >14˚, from northern Scotland to Spain, and ddRADseq was used to reveal the true past of these `feral’ cabbages and their current relationships to each other. Interestingly, despite the huge geographical distances, little genetic differentiation was found. Furthermore, amongst this background homogeneity, there were signals of local adaptation. Overall, these results suggest that wild cabbages in the Atlantic range are an important genetic resource and support a recent expansion from a domesticated source.

42 PopGroup52 Oxford 2019 Talks

Speciation in house mice: investigating the disruptions in gene networks associated with hybrid sterility

Katy Morgan, Leslie Turner

University of Bath, UK

Recently diverged lineages in the early stages of speciation offer opportunities to investigate the mechanisms driving reproductive isolation. Two subspecies of house mouse (Mus musculus musculus and Mus musculus domesticus) diverged in allopatry within the last ~500,000 years before forming a narrow hybrid zone running across Central Europe. While the presence of natural hybrids suggests speciation between the lineages is incomplete, hybrid males commonly show reduced fertility relative to their pure subspecies counterparts. Such reproductive barriers are often caused by incompatabilities between alleles of interacting genes, which have evolved independently in different genetic backgrounds. By constructing and comparing gene networks in high versus low fertility hybrid mice, we aim to identify the disrupted genetic interactions associated with reduced fertility. Future work will compare networks of interacting genes in fertile versus low fertility hybrid mice at specific time points during the spermatogenesis process, to gain a comprehensive mechanistic understanding of the reproductive barrier between differentiating subspecies.

The role of protein architecture in adaptive evolution (S)

Ana Filipa Moutinho, Julien Dutheil

Max Planck Institute for Evolutionary Biology, Germany

The frequency and nature of adaptive mutations is a long-standing focus of the study of molecular evolution. Here, we address the impact of structural architecture among protein coding regions on the rate of adaptive mutations. We used population genetics to study molecular evolution on a fine scale by analysing the impact of genetic variants in the different conformations of protein structure. With this, we aimed to understand how protein biophysics and coding sequence evolution influence fitness and adaptation. By using Drosophila melanogaster and Arabidopsis thaliana population genomics data, we fitted models of distribution of fitness effects and estimated the rate of adaptive amino-acid substitutions both at the protein and amino-acid residue scale, across different categories of gene ontology, chaperone affinity, protein-protein interactions, intrinsic protein disorder and structural motifs. We found that most of the adaptive mutations occur at the surface of proteins and that intrinsic disorder only contributes to the relaxation of selective constraints. Moreover, we observe that the functional class of proteins also plays a role in adaptive evolution, with genes encoding for protein biosynthesis and signalling for their degradation exhibiting the highest values for the rate of adaptive substitutions. Hence, we propose that the higher rate of adaptive mutations at the surface of proteins is driven by new inter-molecular interactions, both at the intra-organism, within protein networks, and at the inter-organisms level, through the coevolution with pathogens.

43 PopGroup52 Oxford 2019 Talks

Seascape genomics of a reef-building ecosystem engineer (Sabellaria alveolata)

Anna Muir, Stanislas Dubois, Rebecca Ross, Louise Firth, Antony Knights, Flavia Nunes

University of Chester, UK Institut Universitaire Européen de la Mer, France

Conservation strategies in the context of climate change rest on the premise that populations can show three responses to avoid fitness loss: evade, evolve or acclimatise. It is important to not only understand neutral gene flow patterns as a measure of dispersal potential, but also adaptive genetic variation between populations as a measure of evolution potential. The honeycomb worm, Sabellaria alveolata, is an ecosystem engineer that constructs biogenic reefs along temperate coastlines, which support high levels of biodiversity, making them of conservation concern. The overall aim of this study was to assess whether the honeycomb worm shows neutral and/or adaptive genetic variation across its range and how this relates to temperature. In particular, we aimed to answer the following questions: 1) Is gene flow restricted between reef sites and can population structure be predicted by ocean circulation modelling; and 2) do S. alveolata show local adaptation and how does this relate to local thermal conditions? We sampled from along the latitudinal range of the species and carried out restriction-site associated DNA sequencing (RADseq). Outlier analyses were carried out to identify loci under balancing or adaptive selection, putatively neutral loci were used to assess population structure in light of ocean circulation modelling, and outlier loci were assessed in relation to local thermal conditions. We identified an isolated population within the species range and linked this to larval dispersal as limited by ocean circulation patterns. Our findings also highlight the importance of having accurate occurrence data when considering connectivity and local adaptation.

The link between genome reduction and pathogenicity in Streptococcus suis

Michael Casey, Jane Charlesworth, Catrin Lloyd, Eric Miller, Gemma Murray, Lucy Weinert, John Welch

University of Cambridge, UK

How and why bacteria evolve to cause disease in their hosts is poorly understood. A common observation is that pathogenic bacteria have smaller genomes and fewer genes than closely related non- or less-pathogenic species. However, it is not known how general this pattern is, or why it holds. Here we test for an association between pathogenicity and genome reduction within Streptococcus suis, a bacterial species that commonly infects pigs, and that appears to have made multiple transitions from commensalism to pathogenicity. An association between genome reduction and the transition to pathogenicity could either indicate adaptive or maladaptive genome streamlining and we try to disentangle these hypotheses by looking at associated genome traits.

44 PopGroup52 Oxford 2019 Talks

Divergence in a quantitative trait is independent of background genetic structure across parallel hybrid zones in Heliconius erato and Heliconius Melpomene

Emma V. Curran, Sean R. Stankowski, Roger K. Butlin, Carolina Pardo-Diaz, Camilo A Salazar, Mauricio Linares, Timothy J. Thurman, Nicola J. Nadeau

Department of Animal and Plant Sciences, University of Sheffield, UK Biology Program, Faculty of Natural Sciences and Mathematics, Universidad del Rosario, Bogota, Colombia Redpath Museum and Department of Biology, McGill University, Montreal, Quebec, Canada Smithsonian Tropical Research Institute, Panamá

Hybrid zones, where phenotypically distinct populations meet and interbreed, can help us to understand the process of divergence and how differences between populations can be maintained in face of gene flow. Trait differences controlled by major effect loci can be maintained by strong divergent selection, while the rest of the genome is homogenised by gene flow. However, for quantitative traits, which are controlled by multiple loci of small effect, stronger barriers to gene flow and higher levels of genome-wide genetic differentiation may be necessary to maintain phenotypic differences. The butterflies Heliconius erato and Heliconius melpomene are distantly related Müllerian mimics that show parallel divergence in wing colour patterns between geographical races across South America. Colour pattern differences between races in both species have been shown to be controlled by a small number of major effect loci that usually show sharp clines across narrow hybrid zones. Here we investigate intraspecific hybrid zones in both of these species where, in addition to a colour pattern change controlled by a single locus, there is also a previously unstudied change in a structurally produced blue iridescent colour, which segregates as a quantitative trait. Despite the difference in genetic architecture of the colour and pattern traits, we find that cline widths are not significantly different, suggesting that the quantitative trait is not more prone to erosion by gene flow. Genotyping-by-sequencing revealed that genomic clines are not concordant with the cline in iridescence, suggesting that differences in quantitative blue iridescence occur independently of background genomic differentiation.

Why is genetic diversity lost more slowly than neutral expectations?

R.A. Nichols, W. Sherwin, C. Holleley, L. Mijangos Araujo

Queen Mary University of London, UK

This talk is motivated by some empirical experimental results that followed the loss of genetic diversity in meta-populations of Drosophila maintained in the lab. I argue that the explanations of the results requires belief in two apparently contradictory propositions: 1) generally the segregation of deleterious recessive alleles in Drosophila populations depletes genetic diversity in adjacent loci (a background selection effect). 2) the same genetic processes lead to sustained genetic diversity at these loci in the experimental populations.

45 PopGroup52 Oxford 2019 Talks

Hybrid sterility in yeast

Jasmine Ono, Ozan Bozdag, David W. Rogers, Ellen McConnell, Arne Nolte, Emre Karakoc, Duncan Greig

University College London, UK

Reproductive isolation, and consequently speciation, builds up from a combination of isolating barriers. We investigated the potential genetic causes of intrinsic reproductive isolation between two closely related species of yeast, Saccharomyces cerevisiae and S. paradoxus. F1 crosses between these species are viable but infertile, with less than 1% of their gametes being viable. Two potential causes for this sterility are chromosome mis- segregation during meiosis and genic incompatibilities. All chromosomes in yeast are essential. Thus, if a gamete is missing a single chromosome, it will be inviable. We show that chromosome mis-segregation is common in this hybrid, and that it can be linked to genome-wide sequence divergence. Genic incompatibilities, where genes that work well in their parental genetic background do not work well together in a hybrid genetic background, may also be present. Build-up of these kinds of incompatibilities is thought to be a common mechanism for speciation, and underlies the Bateson-Dobzhansky-Muller model. We present preliminary evidence of such incompatibilities in this system.

Eukaryote genes are more likely than prokaryote genes to be composites (S)

Yaqing Ou and James McInerney

The University of Manchester, UK

Species evolution is diverse, not only from the processual process but also from the introgressive process. In the latter, non-homologous recombination can result in new genes being formed by combining parts of existing genes, and likewise, it can result in breaking genes up; these genes are termed composite genes. We set out to examine the extent to which genomes from cells, viruses, and plasmids contain composite genes. We identify composite genes when a given gene shows partial homology to at least two unrelated genes. In order to further analyze composite and component genes, we abstracted our genomic data into graphical form. We constructed sequence similarity networks (SSNs) from 1,190,265 genes comprising the genomes of 36 eukaryotes, 56 archaea, 90 bacteria, 79 viruses and 1,614 plasmids. We then identified non-transitive triplets of nodes in this network and explored the homology relationships in these triplets to see if the middle nodes were indeed composite genes. We identified 221,043 genes (18.57%) as being composites of at least two other genes or partial genes. Composite genes were found to be distributed across all kinds of genomes and across all functional COG categories. Interestingly, the presence of composite genes is statistically significantly more likely in eukaryotes rather than prokaryotes.

46 PopGroup52 Oxford 2019 Talks

The origins of anthropogenic adaptation in Silene uniflora

Alexander S.T. Papadopulos, Andrew J. Helmstetter, Daniel Wood, Owen Osborne, Laurence Mason, Edward A. Straw, Luke T. Dunning, Michael Chester, Javier Igea, Joe Parker and Jackie Lighten

Molecular Ecology and Fisheries Genetics Laboratory, Environment Centre Wales, School of Biological Sciences, Bangor University, Bangor, UK

Local adaptation to different habitats within a species can form the initial phase of speciation and gives species the potential to respond to changing environments. Natural selection can drive distantly related organisms to adapt to a specific challenge in the same, or similar, way. To determine how organisms will respond to human-induced environmental change, it is vital to understand the degree to which evolution can be predictable and repeatable. This project examines whether human mining activities have led to the rapid and repeated evolution of heavy metal-tolerant ecotypes of Silene uniflora in the last 250 years. Silene uniflora is a mainly coastal species, but has colonised and adapted to heavy metal-contaminated mine sites in the UK and Ireland. Due to the large distances between mine populations, it is likely that mines have been colonised from distinct coastal populations and metal-tolerance has evolved independently on multiple occasions. Using common garden experiments and ddRAD sequencing, we investigate the extent of zinc and copper tolerance across pairs of coastal and mine populations, determine the level of ecotype differentiation at local and regional scales, and establish whether these geographically disparate, metal-tolerant populations have arisen independently.

47 PopGroup52 Oxford 2019 Talks

Dosage compensation in stick insects

Darren J. Parker, Kamil S. Jaron, Patrick Tran Van, Team Synergia, Nicolas Galtier, Marc Robinson-Rechavi, and Tanja Schwander

University of Lausanne, Switzerland

In species with differentiated sex chromosomes, genes on the X (or Z) will be present in different copy numbers in males and females. This creates a problem since for many genes expression is proportional to copy number. To alleviate this problem, organisms have evolved methods of dosage compensation to equalise expression between the sexes. Most model species show ‘complete dosage compensation’ whereby gene expression on the sex chromosomes is up- or down- regulated dependent on which sex they are present in. However, recent studies using non-model species have shown that the extent of dosage compensation is variable. The explanatory factors that influence the amount of dosage compensation are largely unknown. One proposed factor is the amount of differentiation between the sex chromosomes, with greater dosage compensation evolving as genes on the sex chromosomes diverge. X0 systems represent an extreme point in this continuum, since all genes on the X chromosome will be single copy in males; however, the extent of dosage compensation in X0 systems has been largely unexplored. Here we examine dosage compensation across five X0 species of stick insect in the genus Timema, and four tissue-types. We found evidence for complete dosage compensation in all species and tissues with the exception of the reproductive tract. Average expression of genes on the X chromosome in males was approximately half of that of females in the reproductive tract, however this is likely due to a large excess of female-biased genes in this tissue rather than a lack of dosage compensation.

48 PopGroup52 Oxford 2019 Talks

Tracking selection in time-series population genomic data using ABC random forests

Vitor A. C. Pavinato1, Stéphane de Mita4, Jean-Michel Marin3, Miguel Navascués2

1UMR CBGP, INRA; Université de Montpellier 2UMR CBGP, INRA 3UMR IMAG, Université de Montpellier 4UMR IAM, INRA

Traditional population genetic studies use genotypic or allelic frequency data obtained from several populations sampled at the same time point. However, temporal population genetics data offers a more powerful way of studying complex dynamics, since we can follow allele frequency changes over time in the population. Recent theoretical works have shown that the interaction between the signal of demography and selection can lead to bias in the inference of population size and the false identification of adaptive loci in genome scans. Disentangling the effects of selection and demography is a long-standing difficulty in population genetics, and, in this context, the joint estimation is a necessity, however not yet fully implemented. One potential solution is the co-estimation of neutral and selective parameters using simulation-based methods as Approximation Bayesian Computation (ABC). However, traditional ABC approaches are computationally expensive, and their implementation in complex settings was unrealistic until recently. The introduction of random forests in ABC reduced the computational burden, making it possible to study complex dynamics with few simulations. We propose the use of ABC Random-Forests to implement the joint inference and co-estimate neutral and selective parameters in temporal population genomics datasets. Our results show that the proposed framework can jointly infer demography and selection, allowing to distinguish true demography (census size) from genetic drift (effective population size), as well as estimate the population genetic load (selection parameter).

49 PopGroup52 Oxford 2019 Talks

TBA

Lucy Peters, Susan Johnston, Josephine Pemberton

University of Edinburgh, UK

A central goal in evolutionary biology is to understand how genetic variation is maintained in traits under selection. Determining the genetic architecture of phenotypic variation can help us investigate this question, as this allows us to gain a better understanding of how traits evolve in response to selection. Until recently, researchers have used pedigree data from long-term ecological studies to estimate trait heritability and examine genetic correlations between phenotypes and fitness related traits. Whilst this approach has helped to explain evolutionary change and/or stasis in some cases, it fails to identify genes and genomic regions associated with trait variation, which can elucidate its evolution. Here, we use genomic approaches such as genomic relatedness, genome-wide association studies (GWAS) and genome partitioning to better understand the genetic architecture and evolution of a sexually-selected trait, antler morphology, in a wild population of red deer. We examine ~10 measures of antler morphology, including principle components, from >1000 antlers to determine if these traits are heritable, and how pedigree and genomic estimates compare. Red deer stags have substantial genetic variation associated with antler morphology (h2 ~ 0.4), contradicting theoretical predictions that such traits should have low genetic variation. We then investigate the genetic architecture of the antler measures using genome-wide association and multi-locus partitioning approaches to determine if trait variation is associated with particular genomic regions. This work provides a foundation to understanding the contribution of specific genomic regions to trait variation, and ultimately, how the genetic architecture of these traits co-vary with individual fitness.

50 PopGroup52 Oxford 2019 Talks

Optimal long-read assembly of the red fire ant's genome through parameter-space exploration (S)

Anurag Priyam, Eckart Stolle, Yannick Wurm

Queen Mary University of London, UK

Accurate assembly of genomes is important for measuring and interpreting genetic variation. Accuracy of a de novo genome assembly depends on the repeat composition of the genome, and the length and the error profile of sequenced reads. Compared to short- reads (e.g., Illumina), long-reads (e.g., PacBio, Oxford Nanopore) enable more contiguous and complete genome assemblies. However, long-read assembly is not devoid of problems: mis-joins and assembly of repetitive regions remain a challenge. We used PacBio sequencing to create a better reference genome assembly for the red fire ant, Solenopsis invicta. As a major pest, the species is of huge economic and ecological importance. We sequenced a pool of 21 haploid brothers on PacBio Sequel [V2 chemistry] using five SMRT cells. This resulted in 20 billion sequenced bases, representing 44x genome coverage. To obtain optimal set of contigs we searched parameter-space of the popular long-read assembler, Canu. To compare the assemblies we assessed gene completeness, the extent of mis-joins, and the extent to which duplicated regions were collapsed in the assembly. A total of 45 combinations of parameters were tested. This resulted in an assembly that is 4% more gene complete, has 2-fold less mis-joins and 4- fold better repeat resolution than default parameters. The resulting assembly has 2,616 contigs compared to 69,511 scaffolds in the current reference. Our work provides a solid set of contigs for reconstructing chromosomes of this species and highlights the benefits of optimising assembly parameters.

51 PopGroup52 Oxford 2019 Talks

Drift supports inversions in maintaining population differences after the secondary contact

Marina Rafajlović, Arcadi Navarro, Rui M. Faria

Department of Marine Sciences, University of Gothenburg, Gothenburg, Sweden

When local populations pre-adapted to mutually contrasting environments come into a secondary contact, their genetic differences will be diminished by migration. The dynamics of this process depends on the genetic architecture of the differentiation established prior to the contact, on the strength and type of selection, as well as on the extent of migration between the populations during the contact [1]. However, this tendency may be slowed down if fitness-relevant loci reside in genomic rearrangements, such as chromosomal inversions. For example, in the simplest case of two differentiated populations, fitness- relevant loci may reside in a standard chromosomal arrangement in one population, but in the inverted arrangement in the other population. If gene flux between the alternative arrangements is absent, population differences are either higher (e.g. when ecological trade-offs are present) or maintained for much longer times (e.g. when ecological incompatibilities are at work) than in the absence of inversion polymorphisms [1]. However, existing theory suggests that this advantage of chromosomal inversions in maintaining population differences is significantly reduced when there is just a slight level of gene flux (i.e. ~〖10〗^(-8)) [1]. Notably, these results do not account for the effect of random genetic drift that necessarily operates in all natural populations. To our knowledge, the effect of drift on the role of inversions in maintaining population differences has not been studied to date. Here, we used computer simulations to evaluate the effect of drift on the role of inversions in maintaining population differences after the secondary contact in two models involving ecological incompatibilities, with and without epistasis [2]. Our most important finding is that, in the absence of any mutations, rare gene flux between the alternative arrangements is counteracted by genetic drift, so that population differences are maintained for tens of thousands of generations at the level corresponding to the absence of gene flux. In return, this time scale seems long enough for additional population differences to evolve (e.g. assortative mating, habitat choice, novel genetic incompatibilities) and thereby facilitate the process of speciation towards completion. This conclusion is retained when mutations are added to the model, albeit mutations reduce the advantage of inversion polymorphisms. Finally, we evaluate and discuss how many loci captured in an inversion polymorphism, each having a small fitness effects, influence the role of drift and inversion on the maintenance of genetic differentiation established prior to the secondary contact.

References

[1] Feder, J. L., & Nosil, P. (2009). Chromosomal inversions and species differences: when are genes affecting adaptive divergence and reproductive isolation expected to reside within inversions? Evolution: Int. J. O. Evol. 63, 3061-3075. [2] Rafajlović, M., Navarro, A., and Faria, R.M. Drift supports inversions in maintaining population differences after the secondary contact. (in preparation)

52 PopGroup52 Oxford 2019 Talks

The evolution of human-commensalism in house sparrows

Mark Ravinet, Tore Oleide Elgvin, Cassandra Trier, Marina de la Cámara Peña Sepand Riyahi, Mansour Aliabadian, Andrey Gavrilov and Glenn-Peter Sætre

CEES, University of Oslo, Norway

House sparrows (Passer domesticus) are a hugely successful anthrodependent species; occurring on nearly every continent and well known by most people. Yet despite their ubiquity and familiarity, surprisingly little is known about how this intriguing and popular species came to be. We sought to investigate the evolutionary origins of the House sparrow and to identify the processes involved in its transition to a human-commensal niche. We first used whole-genome resequencing of 17 species, covering nearly the entire Passer genus to construct a phylogeny. This suggests Eurasian sparrows (House, Italian and Spanish) diverged and diversified in Northern Africa and the Middle East. Analysis of a resequencing dataset of >250 individuals from across the Eurasian distribution using coalescent modelling confirms that commensal House sparrows most probably arose in the Middle/Near East and then moved westwards into Europe with the spread of agriculture following the Neolithic revolution. To identify genes and traits involved in adaptation to an anthropocentric niche, we compared phenotypes and genotypes of human-commensal and wild lineages of P. domesticus. 3D analysis of skull morphology suggests more robust skull development and increased brain size in commensal house sparrows. We also identified clear signatures of recent, positive selection in the genome of the commensal house that are absent in wild populations. The strongest selected region encompasses two major candidate genes; COL11A – which regulates craniofacial and skull development and AMY2A which has previously been linked to adaptation to high-starch diets in humans and dogs. Our work examines human-commensalism in an evolutionary framework, identifies phenotypic traits and genomic regions involved in rapid adaptation and ties their evolution to the development of modern human civilization.

53 PopGroup52 Oxford 2019 Talks

The role of sex-biased gene expression in adaptive loss of a male sexual trait (S)

Jack G. Rayner, Sonia Pascoal, Nathan W. Bailey

University of St Andrews, UK

The loss of sexual ornaments is frequently observed across taxa, and pleiotropic effects of such losses provide an opportunity to gain insight into underlying dynamics of sex-biased gene expression and intralocus sexual conflict (IASC). Growing evidence indicates male sexual traits are associated with IASC, despite sex-limited expression, suggesting their loss will affect patterns of sex-biased gene expression in both sexes. We investigated this in a field cricket, Teleogryllus oceanicus, in which an X-linked genotype (flatwing) present in Hawaiian populations feminises males’ wings and eliminates their ability to produce sexually selected songs. Silence protects them from a deadly acoustically-orienting parasitoid, and flatwing evolved rapidly under this natural selection pressure. Our results reveal 1) the flatwing genotype more strongly affects female, rather than male, gene expression at adult stages, with a pattern of demasculinisation that is consistent with relaxed IASC; and 2) adaptive loss of male song appears to have been effected by transcriptomic feminisation in developing male wing tissues, suggesting a capacitating role of genetic variation maintained by balancing selection associated with sex-biased developmental pathways. Our results strengthen the view that sex-limited phenotypic expression offers only partial resolution to intralocus sexual conflict, owing to pleiotropic effects of the loci involved, while maintenance of underlying genetic variation associated with sexual dimorphism provides a substrate for rapid adaptive evolution.

A toad on the road: signatures of hybrid zone movement with 3RAD sequencing (S)

I. van Riemsdijk, J.W. Arntzen, G. Bucciarelli, E. McCarthney-Melstad, P. Scott, B. Shaffer, E. Toffelmier, B. Wielstra

Naturalis Biodiversity Centre, Leiden, The Netherlands Leiden University, Leiden, The Netherlands

Traditionally, hybrid zones are thought to stabilise shortly after secondary contact. Long term hybrid zone movement has not often been demonstrated in nature. In France, a hybrid zone between two common toad species, Bufo bufo and B. spinosus, shows indications of B. bufo moving southwards and displacing B. spinosus. We test this hypothesis of long term hybrid zone movement with different scenarios along the hybrid zone for two transects, one in northwestern France and one in southeastern France, using a variant of Restriction-site Associated DNA sequencing with three restriction enzymes (3RAD). The 3RAD data was filtered to include only species specific SNPs. Combining Bayesian genomic clines with geographic clines helped us identify four types of outlier SNPs: (1) SNPs which easily introgress between both species, (2) SNPs which resist introgression between species, possibly related to barrier genes, (3) SNPs which show a relative high amount of introgression into B. bufo, and (4) SNPs which show a relative high amount of introgression into B. spinosus. We quantify these SNPs and compare the way they behave in both transects (e.g. with admixture linkage disequilibrium). This genome wide dataset allows in-depth analysis of long term hybrid zone movement.

54 PopGroup52 Oxford 2019 Talks

Genome-wide sexually antagonistic polymorphisms reveal longstanding constraints on sexual dimorphism in the fruitfly (S)

Filip Ruzicka, Mark Hill, Ilona Flis, Tanya Pennell, Fiona Ingleby, Kevin Fowler, Ted Morrow, Max Reuter

University College London, UK

Understanding the limits to sex-specific adaptation requires a description of its underlying genetic constraints—i.e., of sexually antagonistic polymorphisms, where different alleles at given loci are favoured by selection in each sex. Previous theory has made a number of predictions regarding the identity and functional properties of sexually antagonistic polymorphisms, and their effects on quantitative and population genetic fitness variation. Yet despite this, little is known about the genetics and evolutionary dynamics of antagonistic polymorphisms. To address this, we perform a genome-wide association study of sexual antagonism across ~200 D. melanogaster hemiclones. We identify ~230 independent genetic clusters of antagonistic SNPs. Contrary to classic theory, we find no evidence that these SNPs are preferentially X-linked. Characterising antagonistic SNPs functionally, we find a large excess of missense variants. We also assess the evolutionary persistence of antagonistic variants by examining extant polymorphism in wild D. melanogaster populations. Remarkably, antagonistic variants are associated with multiple signatures of balancing selection across the D. melanogaster distribution range, indicating widespread and evolutionarily persistent (>10,000 years) genomic constraints. Based on these results, we propose that antagonistic variation accumulates due to constraints on the resolution of sexual conflict over protein coding sequences, thus contributing to the long-term maintenance of heritable fitness variation.

55 PopGroup52 Oxford 2019 Talks

Don’t throw the baby out with the bathwater: a RADiKal new solution to processing RAD data (S)

Camilla Ryan, Diana Bell, Lawrence Percival-Alwyn, Matthew Clark, Ian Barnes, Carl Jones, Cock Van Oosterhout, Bernardo Clavijo

University of East Anglia & Earlham Institute, UK

RAD sequencing has become a popular reduced sequencing strategy in recent years particularly in conservation and ecology. It is important that RADSeq data is processed appropriately so that any results produced can be used with confidence, for example when using genetic data to draw up management plans for endangered species. STACKS, the most popular software for processing RAD sequencing, relies on the data having been cleaned and aligned using external software then employs a number of filters and constraints. At each of these stages true information is being lost which typically results in a vastly reduced number of markers and no way to validate these results. We present RADiKal an alignment free software package for identifying RAD markers using kmers. Raw reads are not subject to filtering but by reducing the sequencing data into kmers of a given length, data from RAD sites can be identified due to their signature which is distinct from errors and repeats. SNPs can easily be called and loci visualised manually to validate the presence of a RAD site.We compare STACKS and RADiKal using RAD data generated from the endangered pink pigeon (Nesoenas mayeri). We compare the number of markers produced and how that impacts subsequent population genetic analysis (e.g. heterozygosity (He), nucleotide diversity (p), genetic differentiation (Fst)). Here we offer an alternate to current data processing which ‘throws the baby out with the bathwater’ and results in a reduced dataset that may alter population genetic inference and resulting advice for management strategies.

56 PopGroup52 Oxford 2019 Talks

Population genetics of human hunter-gatherers: genetic adaptations for social cognition?

Helmut Schaschl, Tobias Göllner, Helmut Lukas

University of Vienna, Austria

We have still a significant gap of knowledge about the genomic diversity of indigenous human populations in particular that of hunter-gatherers despite the fact that about 90% of human history is marked by hunting and gathering as the major form of living. The Maniq are one of the few primarily hunter-gatherers living in the rainforest in Southern Thailand. Together with the Maniq we started a study to analyse their genomic diversity, nutrient diversity and dietary. In a first pilot study, we analysed about 2.6 Mio. SNP markers. We used FST-statistics and performed genome-wide scan for recent positive selection. We found the HLA region, which contains many key genes involved in immunity, is under strong positive selection. However, we detected also outlier loci and loci under positive selection in genomic regions that are involved in brain development and social cognition. For example, a non-synonymous substitution (C-allele) in the oxytocin receptor gene (OXTR) is fixed in the Maniq whereas this allele has a minor allele frequency of <0.02 in the 1000 Genome populations. This is in particular interesting as the neurophysiological effects of oxytocin, which are mediated via its specific receptor molecule the OXTR, are associated with facilitating social communication, affiliative behaviours, and social cognition, but also with neurological conditions such as the autistic spectrum disorder. We thus conclude that genes involved in brain development and social cognition show specific adaption to the hunter-gatherers lifestyle.

Characterizing the genomic signature of polygenic adaptation

Christian Schlötterer

Institut für Populationsgenetik, Vetmeduni Vienna, Austria

For decades, population genetics focused on selective sweeps arising from the spread of beneficial mutations in a population until they ultimately become fixed. The genomic signature of hard sweeps, soft sweeps and incomplete sweeps are well-understood. In contrast, the signatures of polygenic adaptation are less understood. The reason mainly lies in the general assumption of many loci responding by small allele frequency changes, which are difficult to detect. I will discuss how experimental evolution studies in sexual organisms, which include time series genomic data and replicate populations, can detect polygenic selection signatures and distinguish them from selective sweeps of independent loci. Experimental evolution in Drosophila shows that that many large effect alleles contribute to adaptation, but display different trajectories than selective sweeps. Several of the large effect alleles occur at high frequency in the natural founder population. These results demonstrate the potential of experimental evolution to understand the genomic signature of polygenic adaptation. A better understanding of the signatures of polygenic adaptation will be instrumental to develop powerful approaches to identify and characterize selection signatures in natural populations.

57 PopGroup52 Oxford 2019 Talks

Evidence that SIV drove genetic adaptation in natural populations of eastern chimpanzees

Joshua M. Schmidt, Marc de Manuel Montero, Tomas Marques-Bonet, Sergi Castellano and Aida M. Andrés

UCL Genetics Institute/Department of Genetics, Evolution and Environment, UK

All four subspecies of chimpanzees are endangered. Differing in demographic history and inhabiting separate geographical ranges within sub-Saharan Africa, they have likely adapted to different environmental factors. We analysed genome-wide patterns of subspecies differentiation and show that the distribution of highly differentiated SNPs in eastern chimpanzees is uniquely enriched in genic sites in a way that is compatible with recent adaptation, but not with genetic drift and background selection. These sites are enriched in an important set of genes: those that differentiate the immune response to infection by simian immunodeficiency virus (SIV) in natural vs. non-natural host species. Conversely, central chimpanzees exhibit selective sweeps at three cytokine receptors that are paralogs of CCR5 and CXCR4 which are the two major receptors utilized by HIV to enter human cells. Thus, we infer that SIV is eliciting distinctive adaptive responses in these closely related subspecies: in central chimpanzees to limit cell entry, in eastern chimpanzees to modulate the immune response. Since central chimpanzee SIV is the source of the global HIV/AIDS pandemic, understanding the mechanisms that limit pathogenicity of SIV in chimpanzees can broaden our understanding of HIV infection in humans.

58 PopGroup52 Oxford 2019 Talks

The Genomic Landscape of Divergence Across the Speciation Continuum in an Island-colonising Bird (S)

Sendell-Price, A.T., Ruegg, K.C., Anderson, E.C., Quilodrán, C.S., Van Doren, B.M., Underwood, V.L. Coulson, T., Clegg, S.M.

University of Oxford, UK

Understanding the drivers of global diversity patterns requires knowledge of evolutionary dynamics that transform undifferentiated taxa into reproductively isolated species. A goal of genomic-era analyses is to infer the evolutionary dynamics at play during the process of speciation by analysing the distribution of genetic differences across the genomes of diverging forms - also referred to as the genomic landscape of divergence. However, empirical assessments of the genomic landscape under varying evolutionary scenarios are few, limiting the ability to achieve this goal. Here we combine genome-wide sequencing and individual-based simulation models to evaluate the genomic landscape in the silvereye (Zosterops lateralis) - a naturally replicated system that serves as a proxy for the speciation continuum. Using comparisons matched for divergence timeframe and gene flow, we document how genomic patterns accumulate as divergence progresses. In contrast to previous predictions, our results provide limited support for the idea that divergence accumulates around loci under divergent selection or that genomic islands widen with time. While a small number of genomic islands were found in populations diverging with and without gene flow, in only one case were outliers tightly associated with a genomic island. Furthermore, we succeeded in modelling the transition from localised to genome-wide levels of divergence using individual-based simulations that considered only neutral processes. Our results challenge the ubiquity of existing verbal models to explain the progression of genomic differences across the speciation continuum and instead support the idea that divergence both within and outside of islands is important during the speciation process.

Genomic prediction in a wild mammal population

Jon Slate, Bilal Ashraf, Darren Hunter, Phil Ellis, Susan Johnston, Jill Pilkington, Camillo Berenos, Josephine Pemberton

University of Sheffield, UK

A major challenge in evolutionary ecology is the prediction of an individual’s phenotype. In human medicine and in agricultural science, the accessibility of next-generation DNA sequencing has led to the development of ‘genomic prediction’, where phenotypes with a genetic component can be predicted from an individual’s DNA sequence. To date genomic prediction has rarely, if at all, been practised in evolutionary ecology, despite it’s potential to help with practical problems (e.g. dealing with ash dieback) and to understand evolutionary processes (e.g. identifying microevolutionary changes following episodes of selection). Here I will describe how the Soay Sheep long term study can be used to test the accuracy of genomic prediction in a natural system.

59 PopGroup52 Oxford 2019 Talks

Evidence for widespread balancing selection in the human genome

Adam Eyre-Walker, Vivak Soni, Michael Vos

University of Sussex, UK

The role that balancing selection plays in the maintenance of genetic variation remains unknown. In humans, the signature of balancing selection has been detected in various genomic localities, but the overall frequency of balancing selection has yet to be quantified. We present a McDonald-Kreitman type test that compares the number of non-synonymous (sN) and synonymous (sS) polymorphisms that are shared between populations or species, to those that are private (rN and rS respectively) to one of these populations. If some non- synonymous mutations are subject to long-term balancing selection, then we expect (Sn/Ss)/(Rn/Rs) > 1. We can also make a conservative estimate of the proportion of shared non-synonymous mutations that are subject to balancing selection from this statistic. Through simulations we have proven that our statistic has power to detect balancing selection. As tMRCA increases our test gains power due to recombination breaking up linkage between balanced polymorphisms and shared neutral genetic variation. Currently we are using further simulations to understand how this test fares under different demographic scenarios. We have applied this statistic to human population genetic data taken from the 1000 genome project, along with data from Neanderthals, and Chimpanzees. We detect this signal of balancing selection consistently between 1000 genomes continental populations, between humans and Neanderthals, and between humans and Chimpanzees, when low frequency private polymorphisms are excluded.

60 PopGroup52 Oxford 2019 Talks

Comparative Phylogeography in European Hymenopteran Parasitoids (S)

William Walton, Konrad Lohse, Graham Stone

University of Edinburgh, UK

One of the main aims of the field of phylogeography is to understand and predict how geological and climatic events shape the genetic variation within species. These events could cause rapid changes in population size, which are known to have strong effects on genetic diversity. The phylogeography of the Western Palearctic region has been well studied, with several characterised refugia where genetic diversity has been maintained in many species during periods of range contractions in ice ages. However, few studies have explicitly compared the population histories of several European species within a guild in order to determine their responses to geological events and the degree to which these responses are correlated across species. I have used a demographic inference approach based on a blockwise summary of genomic data to determine what population size changes occurred in Iberian refugial populations of seven species of parasitoid wasp during the Quaternary period. I have also used PSMC to corroborate the blockwise results and link them to the history of other refugial populations. I found evidence of dramatic changes in population size in four species, which occurred as separate events, at varying times in relation to the divergence times between Iberian and other populations. This indicates that these size changes occurred as a result of factors specific to the biology or history of individual species, rather than in response to system-wide events.

Co-variation between antibiotic phenotypes is explained by linkage disequilibrium rather than pleiotropy in a zoonotic bacterium

Lucy Weinert, Nazreen Hadjirin, Eric Miller, John Welch

University of Cambridge, UK

Streptococcus suis is a bacterial pathogen of pigs, which also causes disease in people who eat pork. With no vaccine widely available, S. suis is one of the leading causes of blanket antimicrobial use in pigs worldwide. Given the strong selection pressure, S. suis is a good model of antimicrobial resistance. Using the genomes of 627 isolates collected across the world, we screened for susceptibility to 16 different antibiotics, treating the minimum inhibitory concentration as a quantitative trait. We show (1) a good correlation between genotype and phenotype for many classes of antibiotic; (2) that we are not missing many genomic determinants of antimicrobial resistance in this collection; (3) that antimicrobial susceptibility varies systematically between countries, not due to the segregation of different alleles, but due to systematic differences in allele frequencies; (4) that co-variation between antimicrobial susceptibility phenotypes is common and (5) that this co-variation is explained by co-occurrence of antimicrobial resistance alleles, rather than to pleiotropic effects.

61 PopGroup52 Oxford 2019 Talks

Estimation of Individual Inbreeding Coefficients

Bruce Weir1and Jerome Goudet2

1University of Washington, USA 2UNIL, Switzerland

There is continuing interest in using genetic markers to estimate individual-level inbreeding coefficients for a variety of population genetic applications including conservation genetics, association mapping and detection of inbreeding depression. Common estimation methods include those based on heterozygosities, allelic dosages, or the likelihood of observed genotypes, all at multiple loci. We presented a new estimator, based on allelic matching within individuals compared to matching between pairs of individuals, in Genetics 206:2085-2103, 2017. In this talk we provide an evaluation of this estimator, and a comparison with standard estimators, for simulated and empirical SNP data, and we show its performance for inbreeding depression studies. Our estimator seems to perform better than others, and we attribute this to not using sample allele frequencies p as surrogates for unknown allele probabilities. Regardless of sample size, sample allele counts are not binomially distributed whenever the sample contains related individuals, or individuals other than the target individual are inbred. This relatedness and inbreeding cause terms such as 2p(1-p) to not be good estimates of the corresponding parametric function, and this can be a problem for likelihood and moment estimation methods. Our estimates are for within-population, individual-specific inbreeding coefficients. Their average over individuals in a sample provides an estimate of Fis for the sampled population. We stress that estimation of the inbreeding coefficient for an individual depends on the set of individuals used to provide between individual matching comparisons, and that the ranking of individuals by estimated inbreeding should not depend on the comparison set.

Hybridisation and fitness landscapes

John Welch

University of Cambridge, UK

With sexual reproduction, offspring carry a mixture of their parents' genomes, placing allelic variants in new contexts and combinations. Predicting the fitness of these new combinations is one key to understanding phenomena such as hybrid vigour, the evolution of reproductive isolation, adaptive introgression, and the maintenance of hybrid zones. This talk will show how a simple class of fitness landscapes, based on Fisher’s geometric model, can be used to model hybridisation in each of these cases, accounting for the major patterns in the data, and including classical models as special cases. The simplicity of the model makes it especially useful for studying hybridisation in mixed regimes, incorporating both deleterious and beneficial interactions. It also suggests a complementary approach to genomic data from hybrids, which does not rely on identifying small regions with anomalous effects.

62 PopGroup52 Oxford 2019 Talks

Do species traits drive patterns of phylogeography diversity in butterflies? (S)

Joe Middleton Welling, Saad Arif, Tim G. Shreeve

Oxford Brookes University, UK

Butterflies in Europe have been subject to repeated cycles of population expansion and contraction, in response to glacial-interglacial cycles. The signal for this is evident at both a genotypic and phenotypic level. Previous studies have found that whilst this general pattern is present, there remains a large amount of unexplained variation in the amount of population structuring found between individual species. We ask if this variation is best explained by abiotic factors or biotic traits, and over what ecological or evolutionary timescale these operate. For the three putative Mediterranean refugia (Iberia, Italy and the Balkans) we use CO1 variation to calculate within and between population divergences. We used ecological modelling to calculate the landscape resistance between these refugia, both in the present and during the Last Glacial Maxima (LGM) and also gathered species intrinsic traits from published literature. We then used modelling to calculate which traits best explained the population divergences. The best models included some species intrinsic traits, such as number of generations and flight months, and did not include measures of landscape resistance. A large amount of the variance in the data remained unexplained by our best models. We conclude that biotic traits seem to have a key role in explaining the population structuring, but future research needs to explore these issues in more depth.

63 PopGroup52 Oxford 2019 Talks

The co-occurrence and co-exclusion of evolving objects in prokaryotes

Whelan F.J., Rusilowicz M., McInerney J.O.

School of Life Sciences, University of Nottingham, Nottingham, UK

Throughout evolution, evolving objects (domains, genes, operons etc.) have continuously combined, forming new proteins, gene clusters, and genomes. Horizontal gene transfer, particularly among prokaryotes, has facilitated this combinatorial process. Thus, evolving objects that interact positively or synergistically with each other are expected to co-occur more often than by chance; conversely, evolving objects may avoid co-occurrence, indicating an antagonistic or redundant functionality between objects. In this work, we use methods adapted from graph theory to understand patterns of co-occurrence and exclusion in prokaryotes. We have implemented multi-level graph models in which each node (vertex) is a domain, gene, operon, or species connected by an edge (relationship) to another node to display these coincidence relationships. Our method incorporates the phylogenetic distribution and synthenic distances of genes, and we demonstrate how these concepts can be used to identify conserved clusters of vertical and horizontally inherited units of selection. We apply these multi-level graph models to a variety of datasets including prokaryotic pangenomes, a representative set of prokaryotes, and metagenomic sequencing datasets from human-associated microbial communities. We find evidence for genes that significantly co-occur with each other within these datasets; these genetic clusters include objects from characterized biological pathways but also include genes with unknown functions. Further, we identify genes that exclude each other, indicating evolving objects with antagonistic or redundant biological functions. This work represents a different approach to understanding the evolution of prokaryotes and allows us to draw novel hypotheses as to the potential role of these genetic clusters in prokaryote biology.

64 PopGroup52 Oxford 2019 Talks

Hybrid zone movement in crested newts

Ben Wielstra, Terry Burke, Roger Butlin, Aziz Avcı, Nazan Üzüm, Emin Bozkurt, Kurtuluş Olgun, Pim Arntzen

University of Sheffield, UK UCLA, USA Naturalis Biodiversity Center, Netherlands

Speciation typically involves a stage in which species can still exchange genetic material. Interspecific gene flow is facilitated by the hybrid zones that such species establish upon secondary contact. If one member of a hybridizing species pair displaces the other, their hybrid zone would move across the landscape. Such movement has occasionally been observed over years or even decades. This suggests that hybrid zones have the potential to traverse considerable distances over evolutionary time. Yet, the prevalence of such long-term hybrid zone movement is poorly understood. A key prediction of hybrid zone movement is that that the receding species leaves behind a trail of introgressed selectively neutral alleles within the expanding one. We test for such a genomic footprint of hybrid zone movement in two hybrid zones between crested newt species (genus Triturus) that are thought to have shifted position. The strongly asymmetrical and geographically extensive introgression we uncover in the two crested newt cases provide firm support for hybrid zone movement proceeding over considerable time and space.

Interactions between the sexual identity of the nervous system and the social environment mediate lifespan in Drosophila melanogaster

Ewan Flintham, Tomoyo Yoshida, Sophie Smith, Hania J. Pavlou, Stephen F. Goodwin, Pau Carazo, Stuart Wigby

University of Oxford, UK

Sex differences in lifespan are ubiquitous, but the underlying causal factors remain poorly understood. Inter- and intra-sexual social interactions are well known to influence lifespan in many taxa, but it has proved challenging to separate the role of sex-specific behaviours from wider physiological differences between the sexes. To address this problem, we genetically manipulated the sexual identity of the nervous system - and hence sexual behaviour - in Drosophila melanogaster, and measured lifespan under varying social conditions. Consistent with previous studies, masculinisation of the nervous system in females induced male-specific courtship behaviour and aggression, while nervous-system feminisation in males induced male-male courtship and reduced aggression. Control females outlived males, but masculinised female groups displayed male-like lifespans and male-like costs of group living. By varying the mix of control and masculinised females within social groups, we show that male-specific behaviours are costly to recipients, even when received from females. However, consistent with recent findings, our data suggest the courtship expression to be surprisingly low cost. Overall, our study indicates that nervous system-mediated expression of sex-specific behaviour per se - independent of wider physiological differences between the sexes, or the receipt of aggression or courtship - plays a limited role in mediating sex differences in lifespan.

65 PopGroup52 Oxford 2019 Talks

Asexual bdelloid rotifers have coevolved with virulent parasites for fifty million years

Christopher G. Wilson, Antonios Kriezis, Joshua H. T. Potter, Timothy D. Penny, Cuong Q. Tang, Sophie E. Gresham, Constantinos Xenophontos, Tabassum Mujtaba, Isobel Eyres, Timothy G. Barraclough

Imperial College London, UK

The maintenance of sexual reproduction is an enduring enigma. Asexual lineages have major evolutionary advantages over sexual competitors, including a potential doubling of the population growth rate when investment in males ceases. Nevertheless, obligate asexuals account for less than 1% of extant metazoan species, and these are seldom older than 0.1 Ma. According to one leading hypothesis, asexual lineages are short-lived because they adapt too slowly to sustain long-term coevolution with parasites and pathogens. Here, we show that bdelloid rotifers, a diverse Class of microscopic invertebrates famously lacking males, have been coevolving with virulent pathogens for 50 million years. We reconstructed a dated phylogeny for fungi that are exclusive pathogens of bdelloids, using a related fungus fossilised in Cretaceous amber. We estimate that one genus, Rotiferophthora, is 46.4 million years old (95% CI: 37.5-55.3). To test for reciprocal coevolution, we cross-inoculated twelve bdelloid species with twenty Rotiferophthora isolates, and used Bayesian co-phylogenetic analysis to show that infectivity is an interactive function of host and pathogen macroevolutionary history. Because rotifers are transparent, we were able to observe internal resistance mechanisms attacking ingested spores differently, depending on host and pathogen identities. Ancient and sustained coevolution between bdelloids and Rotiferophthora appears to challenge the general theory that animals require canonical sex to keep up with virulent pathogens. Whether theory and counterexample can be reconciled will depend on elucidating the genetic, immunological and ecological factors that moderate coevolution in this new pathosystem.

66 PopGroup52 Oxford 2019 Talks

Evolutionary Encoding and the Ancestry of Everyone

Jerome Kelleher, Yan Wong, Patrick Albers, Wilder Wohns, Gil McVean

Oxford University, UK

Inferring the evolutionary history of the genome is a fundamental problem in evolutionary biology. However, for sexual species, the genomic history, or ancestry, of individuals in a population is confounded by the fact that different regions of the genome have different histories. We have developed a technique that uses genetic variation to infer evolutionary trees at every point in the genome. Our publicly available inference method, called tsinfer, has the potential to scale to every individual within a species, and is of comparable accuracy to other state-of-the-art methods, even outperforming them in some cases, such as in regions under recent selection. It also provides what we term an "evolutionary encoding" for genomic variation data, which we store in the highly succinct "tree sequence" format, suitable for performing rapid, genome-wide evolutionary analyses. We will describe the steps involved in a tsinfer analysis, and illustrate them by reconstructing deep human ancestry from the 1000 Genomes and the Simons Genome Diversity projects; we can then use this to infer ancestry and compress the variation data from a million human chromosomes, obtained from the UK Biobank project. We will also discuss further developments planned for our method, as well as explore ways to analyse and explore the rich pickings available from a million trees of a million tips each.

Ecological Speciation of Senecio Species on Mount Etna, Sicily (S)

Edgar L.Y. Wong, Bruno Nevado, Owen G. Osborne, Alexander S.T. Papadopulos, Simon J. Hiscock, Dmitry A. Filatov

University of Oxford, UK

Senecio aethnensis and Senecio chrysanthemifolius are sister species inhabiting contrasting environments of high (above 2000m) and low altitude (below 1000m), respectively, on Mount Etna, Sicily. They form a hybrid zone at intermediate altitudes between their respective ranges. Previous work demonstrated that speciation of these Senecio species was likely parapatric, with gene flow continuing ever since their divergence around 100-150 thousand years ago, that coincided with the rise of Mount Etna. Despite their recent divergence, some hybrid breakdown was observed in F2 hybrids, indicating that these species have already evolved some degree of postzygotic isolation. This project aims to reveal the evolutionary forces during on-going ecological speciation driven by adaptation to contrasting conditions at high and low altitudes. Using a combination of phenotypic, genetic and genomic data, we first carried out demographic modelling and inferred key demographic features in the speciation of this system. We compared clines of phenotypic traits and genomic markers, and identified markers under diversifying selection, some of which are promising candidates to test for adaptation on the mountain. Selection is also estimated to be strong in the hybrid zone. Future analyses will allow us to further investigate the selective pressures keeping the two species distinct while maintaining a hybrid zone.

67 PopGroup52 Oxford 2019 Talks

Pervasive population genomic consequences of genome duplication in Arabidopsis arenosa

Patrick Monnahan, Filip Kolář, Pierre Baduel, Christian Sailer, Jordan Koch, Robert Horvath, Benjamin Laenen, Roswitha Schmickl, Pirita Paajanen, Gabriela Šrámková, Magdalena Bohutínská, Brian Arnold, Caroline M. Weisman, Karol Marhold, Tanja Slotte, Kirsten Bomblies, and Levi Yant

School of Life Sciences and Future Food Beacon, University of Nottingham, Nottingham, UK

Ploidy-variable species allow direct inference of the effects of chromosome copy number on fundamental evolutionary processes. While an abundance of theoretical work suggests polyploidy should leave distinct population genomic signatures, empirical data remains sparse. We sequenced ~300 individuals from 39 populations of Arabidopsis arenosa, a naturally diploid-autotetraploid species. We find the impacts of polyploidy on population genomic processes are subtle yet pervasive, including reduced efficiency on linked and purifying selection as well as rampant gene flow from diploids. Initial masking of deleterious mutations, higher rates of nucleotide substitution, and interploidy introgression all conspire to shape the evolutionary potential of polyploids.

Mutational Load and Its Effects on the Rise of Beneficial Mutations within-host -- A genetics insight into infectious diseases

Lei Zhao and Chris Illingworth

University of Cambridge, UK

RNA viruses cause a broad range of acute infectious diseases. The evolution of these viruses may be considered on multiple different scales. Both epidemiological and population genetic approaches have been used to study the evolution of viral populations at within a single host, considering phenomena such as the emergence of new strains and drug resistance. Generically, new viral phenotypes are considered in terms of one or a few mutations arising upon a homogenous background. However, a growing body of evidence suggests that mutational load could be an important factor in these processes. Here we describe a novel evolutionary model for viral evolution and combine evolutionary simulations with experimental data to evaluate the effects of mutational load upon the emergence of beneficial alleles in a within-host viral population. We show that mutational load generates a highly stochastic selective background, making the fate of a given mutation less predictable. Further, viruses with a beneficial trait carry more deleterious mutations, decreasing the overall advantage of drug resistance, or of increased adaptation to a novel host. Our approach aims to make progress in bridging the gap between classical models of population genetics and the biology of viral infections.

68 PopGroup52 Oxford 2019 Posters

POSTERS

1(S)

Modelling Ne to describe allele trajectories from genome-wide time series data

Carolina Barata, Carolin Kosiol

Centre for Biological Diversity, University of St Andrews, UK

Studying evolution in the laboratory has allowed researchers to consistently track phenotypic and genotypic changes throughout an experiment. By inducing environmental changes, one can describe how an experimental population responds to any given selective pressure. Evolutionary biologists have tested theoretical predictions by combining such an Experimental Evolution setup with high-throughput sequencing of pooled samples of individuals (Pool-Seq). This Evolve & Resequence (E&R) approach has allowed some genomic patterns of short-term adaptation to emerge. Despite numerous efforts to study allele frequency changes, methods to find targets of selection are still suffering from low power and high false positive rates. We use a non-parametric statistical model that allows testing for selection in genome-wide time series data sets. The beta-binomial Gaussian process (BBGP) regression models the time dependency of an allele’s frequency trajectory, and it accounts for finite sequencing depth associated with Pool-Seq data. However, in most eukaryote E&R experiments, population sizes are small. Therefore, genetic drift plays a major role in the fate of any given allele. This hinders adaptation from standing genetic variation. We extend the BBGP model by adding a covariance function that explicitly accounts for frequency variance generated by genetic drift. Lastly, we will discuss how the extended BBGP-based test for selection can be applied to Drosophila pseudoobscura experimental populations. The experiment was designed to investigate the response to monogamy and elevated polyandry treatments. Furthermore, we use a temporal method developed by Jónás et al (Genetics, 2016) to estimate Ne and describe differences between autosomes and the X chromosome.

69 PopGroup52 Oxford 2019 Posters

2(S)

Proper Treatment of Linkage Disequilibrium in Sequence Data Analysis Significantly Reduces Estimation Error

Stefanie Belohlavy, Nick Barton

Barton Group, IST, Austria

It is often assumed that, because linkage disequilibrium falls off quickly along the genome, it can be ignored in data analysis. However, when many SNP are involved, even weak LD between loci can accumulate to have a large cumulative effect within windows of genome. In this study I demonstrate how estimation error can be severely inflated when LD is ignored, and how estimates can be improved by properly accounting for haplotype structure. I demonstrate the principle in simulations as well as analytically, and illustrate it with data.

3(S)

Characterizing floral morphology in the orchid hybrid Orchis x hybrida and its parents O. militaris and O. purpurea (Orchidaceae)

Bersweden, L., Schatz, B., Clarkson, J., Leitch, A., Fay, M.

Royal Botanic Gardens, Kew and Queen Mary University of London

Hybridization and introgression are widely considered to have a significant impact on plant evolution. In food-deceptive orchids of the genus Orchis pre-mating reproductive barriers are weak, often resulting in the build-up of large hybrid populations between certain species. However, post-mating barriers appear to be much stronger, suggesting a key role in the maintenance of species integrity. Anthropomorphic Orchis species are an excellent system in which to study hybridization because parents and hybrids are regularly found growing in sympatry. Preliminary data suggest that hybrids between Orchis militaris and O. purpurea are fertile and regularly backcross with one or both of their parents. Here, we study the effects of hybridization on floral morphology in French populations of O. militaris and O. purpurea using geometric morphometrics to quantify petal shape.

70 PopGroup52 Oxford 2019 Posters

4(S)

A forward approach to evaluate outlier scans for reproductive barriers

Gertjan Bisschop, Konrad Lohse

University of Edinburgh, UK

Despite the large number of empirical studies that claim to have found "speciation islands", understanding what exactly we can hope to learn about the interplay of gene flow and selection in the speciation process from genomic data remains a major challenge. On the one hand, it has become clear that all that glitters is not gold: it is currently unclear how well genomic scans can distinguish barriers to gene flow from other demographic and selective processes that are unrelated to speciation but may leave similar signatures in the genomic landscape of divergence and diversity. On the other hand, not all that is gold glitters: barrier loci of small effect are difficult or impossible to identify in outliers scans. I make use of recent algorithmic improvements in forward simulations (SLiM3) to generate sequence data under a fully parameterized speciation history and a fully specified architecture of local adaptation. Analysing this data, I ask i) how readily locally adaptive alleles become fixed between diverging species, ii) how easily and over what timescale such barriers can be detected using simple summaries of relative (F_st) and absolute divergence (d_xy) and iii) how both the evolvability and detectability of barriers to gene flow depend on the nature of selection, the genetic architecture and demographic background. This will allow us to focus our attention on specific speciation histories in order to increase our knowledge on the speciation process and reveal what genes and genomic architectures are actually involved.

71 PopGroup52 Oxford 2019 Posters

5

Wing pheromones in Heliconius butterflies: physiology, behavior, and genetics

Kelsey J.R.P. Byers

University of Cambridge

Butterflies in the genus Heliconius (Nymphalidae) have been extensively studied as key examples of speciation, mate choice, and Müllerian mimicry via their bright color patterns. Recently, male chemical signaling via wing pheromones has been demonstrated in Heliconius melpomene and shown to have an effect on female choice. Although the composition of these pheromones is known, which components of the signal are important - and the genetic basis of their production - is unknown. Our understanding of chemical signaling in reproductive isolation in Heliconius is also limited. Using electroantennography, I examined responses of female H. melpomene and H. cydno to natural and synthetic pheromones of both species, as well as to individual dominant compounds. Of the major components of the pheromones, only octadecanal (26% of the H. melpomene pheromone, absent in H. cydno) provoked a significant response in both species. When female H. melpomene were presented with a choice between a control male and one augmented with additional octadecanal, they showed a slight preference for the control male, and also waited twice as long before mating if they chose the octadecanal-augmented male. I am currently analyzing data from 194 backcross individuals to locate QTL for octadecanal production. This work is the first to show physiological responses to male wing pheromones in Heliconius, and argues for their importance in both mate choice and reproductive isolation in combination with other mate choice signals. It also highlights the potential role of octadecanal, a relatively simple pheromone component, in mate choice and reproductive isolation.

72 PopGroup52 Oxford 2019 Posters

6

Do bacterial pathogens have smaller genomes than their non-pathogenic relatives?

Michael Casey, Jane Charlesworth, Catrin Lloyd, Eric Miller, Gemma Murray, Lucy Weinert, John Welch

University of Cambridge, EMBL-EBI, UK

A major outstanding question in prokaryote genetics and evolution is how bacteria evolve to become pathogenic. Much work has focused on identifying genetic variants involved in the phenotype of pathogenesis, and for searching for known ‘virulence factors’ in newly sequenced genomes in order to predict the pathogenicity of novel isolates, with variable success. Here we take a different approach to this problem. It has been observed many times, across all major groups of bacteria, that pathogens tend to have smaller genomes than their non-pathogenic relatives, but this has never been rigorously tested. Here, we carry out a comparative analysis of pairs of pathogen and non-pathogen species from across Eubacteri to test this hypothesis. We also looked at the relationship between genome size and pathogencity in Streptococcus suis, comparing 834 genomes from strains that differ in pathogenicity. Using these datasets we to try to disentangle whether this is to do with bacterial ecology or pathogenicity per se, and try distinguish which evolutionary processes could explain the observed result.

73 PopGroup52 Oxford 2019 Posters

7(S)

Population Structure and Release Calls of Mongolian Toad, Pseudepidalea raddei

Minjee Choe, Siti N. Othman, Ming-Feng Chuang, Amaël Borzée, Zoljargal Purevdorj, Yikweon Jang

Ewha Womans University, South Korea

Release calls of anuran amphibians are proved to be phylogenetically helpful and are used to detect cryptic species. Release calls are a type of vocalization most frequently emitted by males, usually when amplected by conspecific males, and consisting of a series of simple, repeated notes. We analyzed intraspecific variation in the release calls of Pseudepidalea raddei, a toad species widely distributed in Northeast Asian countries, among 4 localities in Mongolia. We examined the level of call variation between males in each call property and the influences of temperature, body size of calling male and habitat type on acoustic features. Because the sampling sites were distant from each other and individuals had different colors and patterns, we expected there might be variation in the call properties. To find out there is genetic variation between P. raddei populations in Mongolia, we collected 129 individuals from 10 localities including the ones where the release calls were recorded. We examined population structure of the species by analyzing 4 mitochondrial DNA regions. The results revealed that Snout-vent length (SVL) strongly influenced rise time and habitat type affected call duration. Bayesian inference tree was separated into two major clades and we used this information to compare call properties between clades and call duration was significantly different. The results may indicate presence of either cryptic species or hybridization. However, adequate geographical description of sampling sites, broader population-level variations and statistical comparisons are required before evolutionary and ecological scenarios for P. raddei can be proposed and tested.

8(S)

Towards phylostratigraphy of bacteria: a Bacillus case

Nina Čorak, Domagoj Kifer, Tomislav Domazet-Lošo

Institut Ruđer Bošković, Bijenička cesta 54, HR-10000, Zagreb, Croatia

Genomic phylostratigraphy is a statistically based method developed with the aim of reconstructing macroevolutionary events. This approach includes mapping the genome of interest onto a consensus phylogeny. Resulting distribution of genes over phylogeny enables one to calculate the relative evolutionary age of transcriptomes or proteomes over organismal ontogeny. Similarly, phylostratigraphic maps of all species covered by consensus phylogeny are the starting point in estimating gains and losses of gene families during the evolutional time. These approaches have been successfully applied in macroevolutionary studies of eukaryotes, but have never been used in bacterial systems. Here we show our attempt to build consensus phylogeny and phylostratigraphic map of Bacillus subtilis in the context of bacterial diversification and to use it for evolutionary analysis of Bacillus biofilm transcriptome and proteome data.

74 PopGroup52 Oxford 2019 Posters

9(S)

Phylogenetic analyses show bat communities in the Baja California peninsula harbour a high diversity of novel cryptic ectoparasite species.

Laura A. Najera-Cortazar, Alex Keen, Kieran Neal, and Simon Goodman

School of Biology, University of Leeds, UK

Characterising ectoparasite diversity is fundamental to studies of host-parasite interactions, evolution and conservation, and also for understanding emerging disease threats for some vector borne pathogens. With more than 1,300 species, bats represent the second most speciose mammalian clade, but their ectoparasite fauna are poorly known for most species. Here we use a phylogenetic approach to characterise ectoparasite taxon identity and diversity for 20 species of bats captured along the Baja California peninsula, by sequencing mitochondrial Cytochrome Oxidase C subunit I and nuclear 18S ribosomal gene fragments. Multiple novel lineages of bat bugs (Cimicidae), flies (Nycteribiidae and Streblidae) and soft ticks (Argasidae), were revealed. Within families, the new linages showed more than 10% sequence divergence, which is consistent with separation at the species level. Both families of bat flies showed host specificity, particularly on Myotis species. We also identified new records in the peninsula of the Streblidae bat flies. One Nycteribiidae bat fly haplotype from Pallid bat (Antrozous pallidus) hosts, was found throughout the peninsula, suggesting potential host migration. Different bug and tick communities were found in the North and South of the peninsula, suggesting environmental factors may play a role in their range boundaries. Overall, parasite diversity was greater at higher latitudes, in warmer sites. This study is the first systematic survey of bat ectoparasites in the Baja California peninsula, and therefore will serve as a first step for understanding how ecological and evolutionary interactions shape parasite community structure along environmental gradients.

75 PopGroup52 Oxford 2019 Posters

10

Mathematical modelling of pangenome evolution

Maria Rosa Domingo-Sananes, James O. McInerney

University of Nottingham, UK

Over the last decade, sequencing of increasing numbers of prokaryotic genomes has demonstrated significant flexibility in the gene content within a species and extensive horizontal gene transfer between species. This has led to the concept of the pangenome: the full set of genes present in a species, which includes core genes, those present in all individuals, and accessory genes, those whose presence varies among individuals. The latter can range from unique genes present in a single sequenced strain to those absent in one or a few strains. There is currently a lot of debate about (i) which factors and forces shape the diversity and size of pangenomes, (ii) how is this gene content diversity maintained, and (iii) whether or not accessory genes are on average advantageous or neutral in different prokaryote populations and species. In order to explore these issues, we developed a model to simulate the evolution of gene content prokaryote populations. We explore the effects of gene number, gene gain and loss rates, population size, fitness coefficient distributions and niche-dependent fitness coefficients.

11

Fixation in highly fecund populations

Matthias Birkner, Iulia Dahmer, Bjarki Eldon

Leibniz Institute for Evolution and Biodiversity Science, Berlin, Germany

We consider the probability of fixation of the fitter type in a haploid highly fecund population with 2 genetic types. We assume sweepstakes mode of genetic reproduction. The offspring law therefore admits a heavy (algebraic) tail. We model selection as a simple viability selection. We show that the probability of fixation of the fitter type can be well approximated by an adaptation of a classical result from branching process theory. We also give an approximation (a convergence in probability result) of the conditional time to fixation, conditioned on the event of fixation. We remark that classical diffusion theory does not apply in our case, since the process describing the frequency of allelic types admits jumps due to the sweepstakes mode of reproduction. Looking backwards in time, our heavy-tailed offspring law gives rise to a special case of multiple-merger coalescent.

76 PopGroup52 Oxford 2019 Posters

12

How rapid is the pea aphid radiation?

Varvara Fazalova, Marc Krasovec, Bruno Nevado

University of Oxford, UK

The pea aphid (Acyrthosiphon pisum) complex is a well-studied adaptive radiation consisting of multiple host-plant races each feeding on a different legume species. Previous studies suggest that the pea aphid complex diversified extremely rapidly, with estimated divergence time of 8,000-16,000 years ago. For distantly related host races, we recently found intrinsic post-mating pre-zygotic isolation, accompanied by strong genomic divergence. This suggests that distantly related host races are at the late stage of speciation. The time calibration of previous studies was based on estimation of an unusually high mutation rate on their maternally transmitted obligatory bacterial endosymbiont Buchnera. We performed mutation accumulation experiment for 12 parthenogenetic lines (24 genomes) of different pea aphid host races during 27 generations. Our estimate of spontaneous mutation rate for the pea aphids falls within range of other insect taxa and suggests much older divergence times in the pea aphid radiation. These results may suggest novel scenarios of diversification in this important model of ecological speciation.

13(S)

The evolution of genetic architecture under divergent selection and gene flow

Ewan Flintham, Timothy Barraclough

Imperial College London, UK

The study of speciation has a long and contentious history in evolutionary biology and a general theoretical framework is still far from complete. It has recently been suggested that some longstanding difficulties in the field arise from overlooking one of the major phases in population divergence; the persistence of conditions generating incipient species - an aspect which could be particularly important where ecological selection drives population structuring. In this study we sought to extend existing theory on the evolution of genetic architecture and population divergence driven by trait differentiation, by using individual- based computer simulations to study the evolution of loci underpinning trait divergence in variable selection regimes in the face of gene flow. Our findings were largely consistent with existing models which suggest that gene flow favours the concentration of genetic effects on to smaller numbers of loci. However, we also found that traits controlled by a greater number of loci and with intermediate variance in allele effects were more effective at promoting divergence over longer time scales in the face of variable conditions, even with gene flow. We also found evidence that complementary gene action and cryptic variation are potentially important unappreciated factors in in divergent adaptation.

77 PopGroup52 Oxford 2019 Posters

14

Using population genomics approaches to understand population dynamics in Brazilian L. infantum and identify virulence factors

Forrester, S1, Jeffares, D1, Carnielli, J. B.T1,2, Costa Silva, V3, Costa, D. L3, Costa, C. H.N3, Mottram, J.C1

1Centre for Immunology and Infection, Department of Biology, University of York, United Kingdom 2Laboratório de Leishmanioses, Núcleo de Doenças Infecciosas, Universidade Federal do Espírito Santo, Vitória-ES, Brazil. 3Laboratório de Pesquisas em Leishmanioses, Instituto de Doenças Tropicais Natan Portella, Universidade Federal do Piauí, Teresina-PI, Brazil.

At present, we are underutilizing advances in whole genome sequencing methodologies to understand the population structure of Leishmania in Brazil, and the benefits these data types can provide in facilitating better treatment of visceral leishmaniasis. Clinical trials for miltefosine treatment, which has a high efficacy in treating Leishmaniasis in India, has resulted in a high rate of treatment failure in Brazil. From 26 Piauí isolates we were able to strongly associate parasite genotype with patient outcome, demonstrating patient outcome is a heritable trait of the parasite. Using a genetic marker of miltefosine resistance, alongside other clinical manifestations, we were able to identify a distinct cline of this marker, and regional distinctions in the parasite genotype across Brazil. In order to understand the parasite genome component of patient treatment outcome, and more fully understand the genomic landscape of Brazil, we have embarked on the first large L. infantum population study using genomic sequencing. The high level of heritability, low genetic diversity and evidence of sexual recombination we have observed in these strains, and the additional genomes we have since collected, suggest that GWAS will facilitate the identification of genotypes under selective pressure and balancing selection with a larger sample size. A localised excess of genetic diversity caused by ‘balancing selection’ is one indicator of a gene involved in host-parasite interaction. Preliminary analysis has already identified robust signatures of balancing selection in multiple regions of the L. infantum genome.

78 PopGroup52 Oxford 2019 Posters

15(S)

Understanding the impact of phylloplane biocontrol agents on insects

Kristina Grenz, Louise Johnson, Alice Mauchline, Rob Jackson

University of Reading and the AHDB, UK

Aphids are a major pest to the agricultural and horticultural industry, causing significant yield losses by directly feeding on the crop and acting as vectors for plant diseases. Concern for the environment, human health and species developing resistance to numerous insecticides, has put growers under increasing pressure to find alternative methods of aphid control. Pseudomonas poae is a gram negative bacteria originally found on cabbage roots that has proven to be naturally virulent against aphids. Previous research found P. poae to be highly specific to aphids at no detrimental effect to the host plant. It possesses two toxicity genes that may explain its virulence to aphids and there is also evidence to suggest it deters aphids from colonizing a crop plant. These traits make it a promising potential biocontrol agent for use in glasshouse systems. We have investigated whether we can improve P. poae and evolve desirable traits for a biological control. By employing experimental evolution, we worked to improve P. poae’s survival on the phylloplane as well as its efficacy at killing aphids. We also investigated whether the bacteria is capable of forming biofilms which may increase its survival on the phylloplane and contribute to microbial protection of the crop. Our research was successful in evolving more virulent isolates of P. poae and biofilm forming isolates. However, there were clear trade offs. The biofilm-passaged isolates had dramatically reduced virulence and loss of motility, where as the aphid-passaged isolates increased in virulence but failed to form biofilms.

79 PopGroup52 Oxford 2019 Posters

16

A numerical solution of the Wright-Fisher stochastic differential equation of population dynamics under natural selection at two linked loci

Zhangyi He, Mark Beaumont, Feng Yu

MRC Toxicology Unit, University of Cambridge, UK

Over the past decade there has been an increasing focus on the application of the Wright- Fisher diffusion to the inference of natural selection from time series data of gene frequencies. A key ingredient for modelling the trajectory of gene frequencies with the Wright-Fisher diffusion is its transition probability density function. Recent advances in DNA preparation and sequencing techniques have made it possible to monitor genomes in great detail over time, which presents an opportunity for investigating natural selection while accounting for genetic recombination. However, most existing approaches for computing the transition probability density function of the Wright-Fisher diffusion are only applicable to one-locus problems. To address two-locus problems, we propose a novel numerical scheme for the Wright-Fisher stochastic differential equation of population dynamics under natural selection at two linked loci. The key innovation of our method is that we reformulate the stochastic differential equation in a closed form that is amenable to simulation, which enables us to avoid boundary issues and reduce computational costs. Our approach can be easily plugged into most existing simulation-based inference methods, which makes it possible to estimate selection coefficients from time serial samples while accounting genetic recombination. We utilise extensive simulation studies to investigate the fate of pairs of linked genes under natural selection and find that in some cases accounting genetic recombination is essential for the inference of natural selection.

80 PopGroup52 Oxford 2019 Posters

17(S)

Genomic imprinting as a window into human language

Thomas J. Hitchcock, Silvia Paracchini, Andy Gardner

School of Biology, University of St Andrews, UK

Humans spend large portions of their time and energy talking to one another. Yet despite the importance of language to our daily lives, the selection pressures that have honed this behaviour remain almost entirely obscure. One key unanswered question is whether this activity is primarily selfish (benefitting the individual) or altruistic (benefitting her social partners). Here we show that the phenomenon of parent-of-origin specific gene expression – or “genomic imprinting” – may provide an answer to this question. First, we point out that an individual’s maternal-origin and paternal-origin genes may have different evolutionary interests regarding her investment into language, and that this intragenomic conflict may drive genomic imprinting. Second, we show that the direction of imprint is expected to depend upon whether investment into language is relatively selfish or altruistic, and these predictions mean that empirical data on gene expression patterns may be used to discriminate between these two possibilities. Third, we use these results to derive predictions concerning the impact of a range of mutations and epimutations on clinical disorders of language and related phenotypes. These predictions provide avenues for future investigation into the evolution of language.

81 PopGroup52 Oxford 2019 Posters

18(S)

Hybridisation and introgression in the Scottish wildcat

Jo Howard-McCombe1, Helen Senn2, Dan Lawson1, Michael Bruford3, Andrew Kitchener4, Mark Beaumont1

1University of Bristol, UK 2Royal Zoological Society of Scotland, UK 3Cardiff University, UK 4National Museums Scotland, UK

The Scottish wildcat (Felis silvestris) is one of the UK’s most endangered mammal species, with as few as 200 individuals left in the wild. Hybridisation with domestic cats is the most serious threat to wildcats in the UK, countering all efforts to prevent decline due to habitat loss and persecution. We aim to determine the timescale and mode of introgression in Scottish wildcats and evaluate the effectiveness of current hybrid tests. Using this information, we aim to better inform management of the captive population and generate a robust model for future wildcat conservation; we will generate whole-genome haplotypic data to address gaps in current understanding and inform ongoing conservation work. Presented here are results from preliminary analysis of wildcat population structure using PCA and fineRADstructure. ddRAD data was generated from 108 individuals (4 domestic cats, 104 putative wildcats). We show that a population of wildcats genetically distant from domestic cats is still present in Scotland, though these individuals are found almost exclusively within the captive breeding program; most wild-living cats sampled were introgressed to some extent. We used the dense ddRAD data to evaluate the accuracy of the existing hybrid tests, both genetic and morphology-based, and found the pelage- scoring test to be a somewhat unreliable predictor of wildcat ancestry. We also present some preliminary investigation into loci potentially under selection in wildcats, using pcadapt.

82 PopGroup52 Oxford 2019 Posters

19(S)

The Genetics and Evolution of Sexual Antagonism: testing for sex-specific fitness associated with allelic variation at the fruitless locus

Michael Jardine

University College London, UK

The sexes have been shaped by selection to achieve different evolutionary ‘goals’ resulting in differences in behaviour, physiology and genetic regulation. Since they share most of their genetic material there is likely to be conflict between them since genetic variation selected for in one sex maybe be detrimental to the other and selected against. This genetic disagreement between the sexes is termed sexual antagonism (SA). Theoretical work has encouraged a host of quantitative genetics studies over the past few years that show that sexually antagonistic genetic variation is widespread and common in many species. However, there has been some difficulty in identifying precisely the genes involved. We have recently produced a list of candidate sexually antagonistic loci for the first time which we aim to study in detail to confirm their SA nature and their role in evolution We have chosen to focus on the candidate gene fruitless in Drosophila melanogaster due to its important role sexual development and stable polymorphism. Several lines of D. melanogaster have been created by introgressing a known allele of fruitless (L or S) into an otherwise identical genetic background. We have performed fitness assays on both sexes from these lines to look for patterns concurrent with a role in SA. Both sexes performed best in lines introgressed with the S allele than flies with the L allele. We therefore find no support for SA at this locus. Future work will look at other potentially important traits, larger scale population experiments and other candidate loci.

83 PopGroup52 Oxford 2019 Posters

20

Comparing genomes and sex chromosomes in dioecious Mercurialis plants

Daniel Jeffries, Jörn Gerchen, John Pannell

University de Lausanne, Switzerland

Sex chromosomes have evolved frequently with the evolution of separate sexes in flowering plants. Many of these transitions have been recent, thus, plants offer a valuable opportunity to study the early stages of sex-chromosome evolution, which typically involves the loss of recombination of the Y (or W) chromosome, and the subsequent degeneration of the non-recombining region. We recently found that the wind-pollinated herb Mercurialis annua has one of the youngest Y chromosomes yet known in plants, with a large non-recombining region that shows only very mild signs of degeneration, including a slight increase in the proportion of non-synonymous mutations and the loss of only a single gene. Surprisingly, although M. annua shares a sex-determining region with its dioecious sister species M. huetti, much of the non-recombining region of its Y chromosome has diverged since the two species split, suggesting that the non- recombining region of M. huetii might be much smaller, and that the large non-recombining region of M. annua has acquired a new stratum very recently by expanded recombination suppression. Here, we examine this hypothesis by comparing the assembly of the M. annua genome with a de novo assembly for M. huetii. We discuss our results in the light of the hypothesised role of both M. annua and M. huetii in the evolution of androdioecy (the co-existence of males and hermaphrodites) in a related allohexaploid lineage via hybridisation and genome duplication.

84 PopGroup52 Oxford 2019 Posters

21

Biophysics and population size constrains speciation in an evolutionary model of developmental system drift

Bhavin S. Khatri, Richard A. Goldstein

Imperial College London, UCL, UK

Developmental system drift is a likely mechanism for the origin of hybrid incompatibilities between closely related species. We examine here the detailed mechanistic basis of hybrid incompatibilities between two allopatric lineages, for a genotype-phenotype map of developmental system drift under stabilising selection, where an organismal phenotype is conserved, but the underlying molecular phenotypes and genotype can drift. This leads to number of emergent phenomenon not obtainable by modelling genotype or phenotype alone. Our results show that: 1) speciation is more rapid at smaller population sizes with a characteristic, Orr-like, power law, but at large population sizes slow, characterised by a sub-diffusive growth law; 2) the molecular phenotypes under weakest selection contribute to the earliest incompatibilities; and 3) pair-wise incompatibilities dominate over higher order, contrary to previous predictions that the latter should dominate. The population size effect we find is consistent with previous results on allopatric divergence of transcription factor-DNA binding, where smaller populations have common ancestors with a larger drift load because genetic drift favours phenotypes which have a larger number of genotypes (higher sequence entropy) over more fit phenotypes which have far fewer genotypes; this means less substitutions are required in either lineage before incompatibilities arise. Overall, our results indicate that biophysics and population size provide a much stronger constraint to speciation than suggested by previous models, and point to a general mechanistic principle of how incompatibilities arise the under stabilising selection for an organismal phenotype.

85 PopGroup52 Oxford 2019 Posters

22(S)

Parallel adaptation to serpentine challenge in Arabidopsis arenosa

Veronika Konecna, Anita Bollmann, Levi Yant, Filip Kolar

Faculty of Science, Charles University, Czech Republic

Adaptation represents key evolutionary mechanism, which allows species and populations to succeed in a variable environment. Some of the outstanding adaptations in plants include adaptations to the hostile serpentine environment. This chemically extreme substrate provides multiple challenges to plant life, for instance, extremely low Ca:Mg ratio, elevated levels of toxic metals and lack of macronutrients. Therefore, serpentine barrens provide a powerful model for studying multi-challenge adaptations. Moreover, the island-like distribution of serpentines can lead to parallel evolution at the level of both genome and phenotype. We have chosen Arabidopsis arenosa as a promising model for studying substrate adaptive evolution in serpentines using state-of-the-art population genomic methods including genome scanning for selection candidates and using high- throughput elemental approach for studying plant ionome. By sampling and genotyping four pairs of serpentine and close non-serpentine populations from the Czech Republic and Austria, we ask if serpentine populations of A. arenosa from central Europe are the result of parallel evolution. First results have shown multiple colonization events of serpentine barrens as well as several common candidate genes for selection among serpentine populations were identified. Furthermore, the candidate genes have relevant adaptive functions to serpentines such as in dehydration tolerance and ion homeostasis traits. We found a signiture for convergent evolution with A. lyrata by comparing the outlier genes. Moreover, we will ask if adaptive introgression from this closely related species played a role in serpentine adaptation.

86 PopGroup52 Oxford 2019 Posters

23

New features for polymorphism-aware phylogenetic models

Dominik Schrempf, Bui Quang Minh, Rui Borges, Carolin Kosiol

University of St Andrews, UK Vetmeduni Vienna, Austria

We have introduced a polymorphism-aware phylogenetic model (PoMo, De Maio et al., 2013), which overcomes computational limitations of multispecies coalescent methods with the abundant population data now available for many species. PoMo scales well with the increasing amount of sequence data while accounting for present and ancestral polymorphisms. PoMo circumvents handling of gene trees and directly infers species trees from allele frequency data. The PoMo approach also allows the modelling of allelic selection. We prove that the Markov process underlying our PoMo-Selection model is reversible. The derivation of the stationary distribution broadens the PoMo approach to further implementations within Bayesian and Maximum Likelihood (ML) frameworks that use site frequency spectra as input data. Furthermore, we extend the ML implementation of PoMo in the software package IQ- TREE (Nguyen et al., 2015). We integrate a search for the statistically best-fit mutation model, the ability to infer mutation rate variation across sites, and assessment of branch support values. We exemplify an analysis of a hundred species with ten haploid individuals each, showing that PoMo can perform inference on large data sets. While PoMo is more accurate than approaches using standard substitution models, it is almost as fast. The new features consolidate the value of PoMo for phylogenetic analyses with population data.

87 PopGroup52 Oxford 2019 Posters

24(S)

Post-glacial and modern demographic history of barn owls (Tyto alba) in the British Isles revealed by whole-genome resequencing

Ana Paula Machado, John Lusby, Guillaume Dumont, Alexandre Roulin, Jérôme Goudet

Department of Ecology and Evolution, University of Lausanne, Switzerland

Since Darwin’s first visit to the Galapagos in 1835, islands and the atypical organisms that inhabit them have fascinated biologists and contributed greatly to the development of evolutionary theory. The recent advances in sequencing technology have boosted research in this field by generating large scale data that allows us to address new questions. Specifically, one can now attempt to disentangle the roles of demographic history and natural selection in generating intraspecific genetic and phenotypic diversity of island populations. In continental Europe, barn owls (Tyto alba) display a plumage colouration cline: owls in the South are mostly white coloured whereas in the North they are dark-reddish. Unlike the neighbouring populations in North-western Europe, barn owls in the British Isles are predominantly white. The reasons for such disparity in coloration have thus far eluded our understanding. Since genetic variation in Europe is relatively shallow, low-representation markers available until now did not allow to address these questions, thus we produced whole-genome resequencing data for 77 individuals in western Europe and the British Isles. We intend to unravel the colonization history of the latter by barn owls in order to understand whether neutral evolutionary mechanisms could be responsible for these populations’ white coloration or whether adaptive forces play a role. Here, we present preliminary results.

88 PopGroup52 Oxford 2019 Posters

25(S)

Landscape of genomic selection across island bird populations

Claudia A. Martin, David S. Richardson, Brent C. Emerson, Claire Armstrong, Lewis G. Spurgin

School of Biological Sciences, University of East Anglia. Instituto de Productos Naturales y Agrobiología, CSIC, Tenerife.

Identifying ecologically important traits and establishing the selective pressures on these traits in the wild, is central to understanding the mechanisms behind evolutionary change. To measure evolutionary adaptation in wild populations we must identify phenotypes under selection, understand the genetic basis of those phenotypes and identify potential drivers of selection. Studying island populations of the Berthelot’s pipit (Anthus berthelotii) across the Canaries, Maderian and Selvagens archipelagos has given us in depth understanding of the colonisation history and dispersal of these populations as well as understanding of how environmental pressures vary across these landscapes. Building on our previous understanding, we now investigate how selection operates at different geographic scales across the 13 island populations. Using RAD-sequenced markers across the genome, we first apply reverse genetics approaches to identify signatures of selection within the archipelagos. We then follow up on these genomic regions, linking gene function to genotypic and morphological variation across all island populations. These findings help us to develop an understanding of the landscape of genomic divergence and selection across island populations of the pipit across different habitats and geography.

89 PopGroup52 Oxford 2019 Posters

26(S)

Local ancestry and functional genomics of trypanotolerant and trypanosusceptible admixed African cattle breeds

Gillian P. McHugo1, Grace M. O'Gorman2, Emmeline W. Hill1, David E. MacHugh1,3

1UCD School of Agriculture and Food Science, University College Dublin, Belfield, D04 V1W8, Ireland 2National Office of Animal Health Ltd., Enfield, EN2 7HF, United Kingdom 3UCD Conway Institute of Biomolecular and Biomedical Research, University College Dublin, Belfield, D04 V1W8, Ireland

African cattle represent a complex mosaic of Bos taurus (taurine) and Bos indicus (zebu) with most breeds containing varying levels of taurine-zebu admixture. The two types of cattle diverged at least 500,000 years ago and significant genomic differences have accumulated since that time. One important evolutionary adaptation in certain African taurine populations is a genetically determined tolerance to infection by trypanosome parasites (Trypanosoma spp.), which are transmitted by infected tsetse flies (Glossina spp.) and cause African animal trypanosomiasis (AAT) disease. The annual financial burden of AAT is approximately $4.5 billion and AAT is one of the largest constraints to livestock rearing in the areas of sub-Saharan Africa with significant tsetse densities. The West African taurine N’Dama breed is trypanotolerant; they have the ability to control parasite loads and to limit disease pathology compared to trypanosusceptible zebu breeds. However, zebu or zebu-taurine hybrid animals are generally larger, produce higher milk yields and are therefore favoured by many farmers. Using local ancestry analysis of genome-wide high-density SNP data, we have examined hybrid West African cattle populations to study sub-chromosomal admixture. These results can also be integrated with gene expression data from bovine trypanosome infection studies to further explore the functional biology of differentially introgressed genomic regions.

90 PopGroup52 Oxford 2019 Posters

27

Barriers and drivers of gene flow across ploidy barrier in wild Arabidopsis

Morgan E.J., Lučanová, L., Kolář F.

Charles University in Prague, Czech Republic

Polyploidy, caused by whole genome duplication, is considered a strong reproductive barrier which can lead to population divergence and, eventually, speciation. Recent advances, however, show that ploidy barrier is often permeable and interploidy gene flow regularly occur. The diploid-autopolyploid complex of Arabidopsis arenosa provides a suitable system for addressing triggers and evolutionary consequences of gene flow, as several stable ploidy contact zones across Europe have been identified. At these contact zones, the two cytotypes are separated in space, although reproductive isolation is imperfect. We use a combination of field surveys, crossing experiments and flow cytometry to explore potential environmental and developmental reasons for the isolation. We assess levels of gene flow and how this varies in the natural system, as well as the potential for admixture among the different cytotypes in the field and in controlled crossing experiments. Using whole genome resequencing we aim to describe the genomic landscape of inter-ploidy admixture and address its evolutionary consequences.

91 PopGroup52 Oxford 2019 Posters

28(S)

Gene fusion events in Metazoa – patterns of emergence and their potential as phylogenetic markers

Peter Mulhair, Raymond J. Moran, Chris J. Creevey, Bede Constantinides, Ian M. Carr, James McInerney, Davide Pisani, Mary J. O’Connell

University of Leeds, UK University of Nottingham, UK

Gene fusion events occur when two or more distinct protein-coding regions are physically remodelled and combined to form a single open reading frame. These rare genomic events are a potential source of rapid protein evolution and may have contributed to the evolution of complex traits within animals. Using a dataset of 1.2 million protein coding genes from 63 metazoan taxa and applying a sequence similarity network-based approach we identify a total of 18,519 gene fusion families distributed across the metazoa. Metadata analysis of transcriptome data confirmed whether the break-points of gene fusion events were expressed in a single open reading frame. Following alignment and comparison of the gene fusions with their parent genes, we mapped the patterns of emergence and found large numbers of families emerging at specific nodes across the metazoa including Bilateria, Gnathostomata and Mammalia. The emergence of the gene fusion families coincident with nodes of major transition in the animal tree indicates the importance of these remodeling events for animal protein evolution, and suggests that these events may contain a strong phylogenetic signal. To assess the phylogenetic properties of gene fusions we assessed whether the pattern of presence/absence of gene fusion families recapitulated uncontroversial, well-resolved nodes within the animal phylogeny. Our results indicate that gene fusion events may play a role in the evolution of complex traits in animals and have the potential as novel phylogenetic markers for animal phylogeny.

92 PopGroup52 Oxford 2019 Posters

29

Gene regulatory network evolution in East African lake cichlids

Will Nash, Tarang Mehta, Padhmanand Sudhakar, Tamas Korcsmaros, Federica Di Palma, Wilfried Haerty

The Earlham Institute, UK

The East African lake cichlids represent the most diverse radiation of extant vertebrates. This radiation spans a wide range of environmental conditions, presenting a wide array of unique ecological niches. The adaptation seen across the radiation is equally diverse and is represented in many facets of cichlid biology – form rapid and unique change in feeding strategy to extreme variation in mating system. As this diverse species group has begun to be known at a genomic level, key examples of regulatory evolution between different ecosystems have been shown to underlie the rapid adaptation to both lakes and rivers in East Africa. Utilising the genomes of five cichlid species from across different environments, both riverine and lacustrine, we have described a core regulatory network based on coexpression modules drawn from RNA-seq data. Evolution of this network across the sequenced radiation is linked to differentiation in key traits such as vision. To further understand this evolution, we have extrapolated transcription factor : target gene : transcription factor binding site (TFBS) relationships from existing CHIP-seq datasets on both Human and Mouse to each fish genome using a methodology based on orthology and sequence similarity. Here, we investigate and compare the impact of genome variation on the cis-regulatory regions within the 5 cichlid species reporting gains and losses of TFBs correlating with environmental adaptations. We extended this analysis to a set of genomes from 73 Lake Malawi Cichlids species which has recently been made publicly available. Given the young age of the radiation in Lake Malawi, and the low level of sequence divergence between species pairs, our aim was to further understand regulatory evolution across this diverse radiation. We hypothesised that evolution of both TF and TFBS would be linked to adaptive evolution across this diverse lineage.

93 PopGroup52 Oxford 2019 Posters

30

Population genomics of bdelloid rotifers

Reuben W. Nowell, Chris G. Wilson, Tim G. Barraclough

Imperial College London, UK

The bdelloid rotifers are a class of microscopic invertebrates that have been evolving for ~50 million years apparently without sex or meiosis. Such ancient and obligate asexuality runs counter to current hypotheses that explain the evolutionary advantages of sex and recombination, leading to the description of bdelloids as “evolutionary scandals”. We have recently published the genomes of three species of bdelloid rotifer, and showed that their genome architecture might be more similar to that of sexual taxa than was previously thought. Specifically, we did not find any evidence of widespread collinear breaks or genomic palindromes that would otherwise render the genome incompatible with conventional meiosis. We did, however, recover a strong signal for some other unusual genome characteristics known in bdelloids: a low number of known transposable elements (TEs) and a remarkably high number of genes of “foreign” origin, presumably acquired via horizontal gene transfer (HGT). Here, we expand these results using medium-coverage genome resequencing of five individual rotifers from two closely related bdelloid species, collected at varying geographic distances. We use these data to characterise patterns of TE and HGT presence and absence among individual genomes, and quantify TE insertions/deletions and recent HGT gains both within and between species. Our results shed further light on the evolution of bdelloid rotifers.

94 PopGroup52 Oxford 2019 Posters

31(S)

Molecular adaption of an ice-age survivor - the Saiga antelope

David T. Orr, Ian M. Carr, Simon J. Goodman, and Mary J. O’Connell

University of Leeds, UK University of Nottingham, UK

A key challenge in evolutionary genomics is to understand the genetic basis of adaptation and evolutionary novelty. Saiga antelope (Saiga tatarica), are native to the steppe of Central Asia, have received extensive attention in recent years due to mass mortality events. In 2015, Saiga populations in Kazakhstan were subjected to a mortality event in which 200,000 individuals died in around 20 days. The proximate cause of death has been established to be through the pathogen Pasteurella multocida. P. multocida is a harmless commensal bacterium in Saiga, residing in the respiratory tract, but abnormally high temperatures and humidity contributed to a weakened immune system, which in combination with viral populations spreading into the blood stream, caused haemorrhagic septicaemia. Historically, Saiga are susceptible to these large-scale mortality events and as a consequence have adapted some interesting features to expand populations rapidly, such as frequent twinning and early maturation of females. Here we use transcriptome data, generated by our group, from an individual of Saiga in comparison with other publicly available vertebrate genome assemblies. We sought to assess the unique selective pressure heterogeneity in protein coding regions of Saiga in comparison to its close ungulate relatives. Using substitution rates (Dn/Ds) within single gene ortholog families we detect patterns consistent with the expectation of positive selection. This large-scale selective pressure is performed using the VESPA software package developed by our group. The functions of genes with signatures of positive selection unique to the Saiga lineage are described here and contribute to the unique phenotypic traits observed in modern day Saiga.

95 PopGroup52 Oxford 2019 Posters

32

The genomic basis for sympatric speciation in Lord Howe Island palms

Owen G. Osborne, Adam Ciezarek, Trevor Wilson, Darren Crayn, Ian Hutton, William J. Baker, Colin G.N. Turnbull, Vincent Savolainen

Imperial College London, UK

Along with crater lake cichlids, Howea palms represent one of the most clear-cut cases of sympatric speciation. The sister species H. belmoreana and H. forsteriana are endemic to Lord Howe Island, Australia, where they have overlapping distributions. The species are reproductively isolated by flowering time differences, although occasional hybrids are produced. The island is only 12 km at its longest diameter, making geographical separation during speciation unlikely. However, genomic-scale analysis to identify external secondary gene flow has so far been lacking. Furthermore, the mechanism of speciation in Howea remains unclear. We analysed a large RNA-seq dataset from Howea and their closest mainland relatives, as well as detailed soil chemistry data. Coalescent analyses show no evidence of introgression from mainland taxa, bolstering the case that the species diverged without geographic separation. Expression analysis, population genetic and phylogenetic tests of selection, identified several candidate ‘speciation genes’. The identity of these genes support mechanisms by which the species may have switched from initial local adaptation to the reproductive isolation seen today. Soil chemistry affects the expression of 18 genes involved in flowering time regulation, indicating that soil-driven flowering time plasticity could have initially isolated the species. Furthermore, seven candidate genes under selection are pleiotropically involved in both soil adaptation and flowering time regulation. Thus, local adaptation to the varying soil types of the island may have had a knock-on effect on flowering time, ultimately driving the evolution of reproductive isolation.

96 PopGroup52 Oxford 2019 Posters

33(S)

Phylogenetic analysis of the critically endangered Grenada dove (Leptotila wellsi)

Catherine Peters, Ian McDowall, Howard Nelson, Bonnie Rusk and Anna Muir

University of Chester, UK

The Grenada dove (Leptotila wellsi) is a critically endangered species endemic to the island of Grenada. It exists in three seemingly isolated populations across the island. As an endemic on a small island with a small and decreasing population size this species is at risk from genetic deterioration. Particular risks include inbreeding depression and accelerated loss of genetic diversity resulting from ongoing habitat fragmentation and population isolation. The phylogenetic placement of the Grenada Dove among Columbidae is currently unknown. In this study we conducted a genetic assessment from non-invasively collected feather and egg type specimens to explore the phylogenetic relationship of this species with other columbids. Mitochondrial DNA (mtDNA) markers, NADH dehydrogenase subunit 2 (ND2) and cytochrome oxidase I (COI) barcoding region, and the nuclear marker, Beta-fibrinogen (FGB) were used to conduct a neighbour joining tree and assess genetic differentiation between species. We have predicted the first phylogenetic placement of the Grenada dove within the genus Leptotila on the basis of analysis of partial regions of mitochondrial and nuclear sequences. Future work will include population structure analysis using non-invasively collected samples from the last three known Grenada dove sites. This information will be used to advise conservation programmes in regards to population management of this species.

97 PopGroup52 Oxford 2019 Posters

34

Why a Y? The Y chromosome drives earlier male development in the common frog (Rana temporaria)

Barret C. Phillips, Nicolas Rodrigues, Anouk Athanasiades, Roberto Sermier, Julien Leuenberger, Guillaume Fabre, Nicolas Perrin

University of Lausanne, Switzerland

Polymorphisms in sex determination systems can produce insights into the link between genotype and gonadal development. In the early twentieth century, Emil Witschi described ‘sex races’ that differed in their degree of gonadal differentiation at metamorphosis in the common frog and suggested their distributions might be linked to both genotype and climate. More recent work has established that the presence and degree of differentiation of sex chromosomes varies both within and between Rana temporaria populations. We investigated the link between sex chromosomes and sex races by raising families from 14 populations. Contrary to initial expectations, the degree of sex chromosome differentiation does not influence an individual’s developmental trajectory—rather, the rate of testis differentiation is determined by the presence or absence of a male genotype at the candidate sex-determining gene, dmrt1. That is, in families where the father was XY, males were more likely to be present at metamorphosis, whereas in families with an XX father, most or all individuals had developing ovaries or undifferentiated gonads at metamorphosis. However, this difference between XX and XY families disappears at the froglet stage, where we find no significant differences in family sex ratios. Thus, in the common frog, the presence of a Y chromosome drives earlier male development, and the sex races correspond to the proportion of XY males in the source populations. Finally, we also directly confirm the presence of truly XX males in Rana temporaria, and thus that sex determination in the common frog is only partly genetic.

98 PopGroup52 Oxford 2019 Posters

35

Extreme and maladaptive GC bias in the evolution of the fat sand rat genome

Rodrigo Pracana, Peter W.H. Holland

Department of Zoology, University of Oxford, UK

Biases in mutation and recombination can cause the fixation of non-adaptive and maladaptive alleles in populations. For instance, deleterious mutations can spread in the population by processes such as GC-biased gene conversion or as a result of low recombination rates. The fat sand rat (Psammommys obesus, a species of gerbil) has recently been shown to carry a ~10 Mb region of the genome with approximately 80 genes that have an extremely high GC content. One of the genes in this region, PDX1, is a homeobox gene essential for the development and function of the pancreas. Despite being 100% conserved across other mammals, the homeodomain of the fat sand rat PDX1 has 15 amino-acid differences relative to its sequence in the mouse. This suggests that a maladaptive process has caused the high GC of genes in the region. However, it is currently unclear what the molecular mechanism underpinning this bias is. Here, we aim to answer the following questions. (1) Do all genes in the region show such an accumulation of AT to GC nonsynonymous mutations? (2) Does the gerbil genome carry other similar regions of high GC? (3) Do other mammal species carry such regions? Answering these questions will help us understand whether maladaptive increases in GC are common across the mammalian lineage and whether they affect chromosome evolution.

99 PopGroup52 Oxford 2019 Posters

36

The effects of coastal infrastructure on the dispersal and gene flow of native and non-native intertidal marine invertebrates

Melanie Prentice, Joe Ironside, Paul Shaw

Aberystwyth University, UK

Artificial sea defences including sea walls, breakwaters and groynes already populate many areas of coast around the Irish Sea, and are likely to become increasingly prevalent as incidences of adverse weather events and sea level rise threaten coastal communities and infrastructure. Constructed of stone, concrete or metal, artificial structures tend to be colonised by organisms adapted to life on rocky shores, but are often placed in areas of soft sediment since these tend to be at greatest risk of erosion. Further, as they usually provide poor habitat for native species, artificial structures are often preferentially colonised by non-native species, particularly when associated with ports, marinas or aquaculture facilities. Ecostructure is a project seeking to understand the ecological and evolutionary impacts of artificial structures on intertidal marine organisms. This poster will outline the population genetic and genomic objectives within the Ecostructure project. First, we aim to assess the population genetic structure of native and non-native marine invertebrates in artificial and natural habitats, in order to understand the role of these structures in facilitating or impeding gene flow in several Irish Sea species. Specifically, we are investigating the potential for artificial structures to act as stepping-stones in the dispersal and gene flow of native rocky shore organisms across areas of unsuitable habitat. If colonised successfully, these stepping-stones may enable some coastal species to extend their ranges northwards to track their optimal environmental niches. We further aim to identify adaptive genetic markers that may provide adaptations for colonising artificial structures. This is especially interesting in the case of the invasive species Didemnum vexillum, which shows variable patterns of invasiveness and colony morphology throughout the Irish Sea.

100 PopGroup52 Oxford 2019 Posters

38

Slightly More Efficient Strategies for Calculating Block-wise Likelihoods Under The Coalescent

Derek Setter, Konrad Lohse

University of Edinburgh, Institute of Evolutionary Biology, UK

The mutation configurations found in small non-recombining blocks of sequence data capture short-range linkage information and provide a more-complete picture of the local genealogical history than do the SFS and other simpler measures of genetic variation. The frequency spectrum of such blocks (bSFS) is a highly informative summary of genetic variation useful for inferring the combined demographic and selective history both within populations and between closely-related species. The likelihoods of the bSFS can be calculated directly from the generating function for the inter-event times in the coalescent framework and has proved useful for analysing data under purely demographic models. However, these calculations are limited to small sample sizes, beyond which, the number of terms in the generating function increases beyond computational feasibility. This poses a particular challenge for analysing models of positive selection, in which, high power to detect the signature of selection in the SFS requires relatively large sample sizes. To ameliorate this limitation, we introduce a directed-graph representation for the possible genealogical histories as well as a recursive function which, given a mutation configuration, efficiently distributes the observed mutants across their possible locations in the genealogy. We show that his new procedure calculates bSFS likelihoods much more quickly for non-small sample sizes and is capable of handling sample sizes otherwise too large for direct calculation from the generating function.

101 PopGroup52 Oxford 2019 Posters

39(S)

Reduced genetic diversity results in highly parallel selection signatures during experimental evolution

Christian Schlötterer, Viola Nolte, Thapasya Vijayan

Veterinary Medicine University Vienna, Austria

Experimental evolution (EE) is a powerful approach to study adaptation under controlled conditions. Combined with sequencing of pooled individuals, EE can be used to trace the allele frequency trajectory through time, an approach called evolve & re-sequence (E&R). Previous E&R studies in Drosophila aimed to use a large number of founder chromosomes for a better mapping resolution. Nevertheless, most studies observed a large number of selection targets which raises the possibility that multiple selected loci are located on selected haplotype blocks. Given that many selection targets are segregating in natural Drosophila populations, we dramatically reduced the variation in the founder population. Using a mixture of two inbred lines, Oregon-R and Samarkand, the populations were subjected to a stressful high temperature environment of 29°C. Tracing the allele frequency changes through across 20 generations in 20 replicates, we observed a pattern of strikingly parallel adaptation. Even after 100 generations the pattern remained highly repeatable across three replicates. Contrasting the high temperature selection regime to a benign environment (18°C), we find evidence for temperature adaptation as well as laboratory adaptation, shared between both environments. We hypothesize that most laboratory adaptation signatures reflect the purging of deleterious alleles. Interestingly, the Samarkand genotype seems to carry more of these deleterious alleles than Oregon. We will use computer simulations to evaluate whether the highly parallel selection signatures can be explained by strong selection operating on a few selection targets or whether a polygenic model provides a better fit to the data.

102 PopGroup52 Oxford 2019 Posters

40(S)

The dimensionality of selection in speciation: an experimental evolution approach

N. White, I. Eyres, M. Brockhurst, R. Snook, A. Beckerman & R. Butlin

University of Sheffield, UK

A finite quantity of divergent selection can be applied either very strongly to a single axis or spread more thinly over multiple different axes. This ‘dimensionality’ of selection can be conceptualised as the number of traits or genetic loci under selection, and variations in dimensionality are predicted to have a variety of consequences for the initiation and progression of speciation. Strong selection along a single axis may rapidly generate strong local adaptation, but not lead to widespread divergence across other isolating traits and genomic regions. Conversely, weaker per-locus selection spread over many axes may be more likely to produce reproductive isolation and wider genomic differentiation, although it may take longer to overcome gene flow, if indeed at all. To test these predictions, we will conduct a long-term evolution experiment with an ‘evolve and re-sequence’ approach, using a species of monogonont rotifer, Brachionus plicatilis. Populations will be exposed to divergent selection along either one or three axes of selection. All selection pressures have been optimised in pilot experiments and the evolution experiment is set to begin presently.

103 PopGroup52 Oxford 2019 Posters

41(S)

Field Cricket Genome Reveals the Signature of Recent Adaptive Trait Loss

Sonia Pascoal, Judith E. Risse, Xiao Zhang, Mark Blaxter, Timothee Cezard, Richard J. Challis, Karim Gharbi, John Hunt, Sujai Kumar, Emma Langan, Jack G. Rayner, Michael G. Ritchie, Urmi Trivedi, Nathan W. Bailey

School of Biology, University of St Andrews, UK

What happens to a genome when an adaptive mutation invades? In natural systems, it is challenging to reconstruct genomic dynamics during the very earliest stages of evolutionary adaptation because it mostly has occurred in the distant past and must be inferred indirectly. We overcame this challenge by capitalizing on the abrupt appearance and spread of a Mendelian variant (flatwing) which silences male crickets (Teleogryllus oceanicus) in Hawaii. Males are attacked by fatal, acoustically-orienting endoparasitoid flies, and flatwing arose and rapidly spread through a population on Kauai in less than 20 generations under this pressure. We sequenced and annotated the T. oceanicus draft genome, constructed a high-density linkage map, mapped quantitative trait loci (QTL), and compared gene expression between flatwing and normal-wing embryos. We localised flatwing on the X chromosome, in a region that coincides with loci associated with chemical pheromone differences between the two male morphs. DE genes associated with the flatwing genotype are also widespread across the whole genome. Our findings illustrate that the genomic impacts caused by recent adaptation are widespread and consequential, exerting effects at the levels of genome architecture, gene expression, and phenotypes, and they add to our understanding of the role of selective sweeps and pleiotropy during the initial stages of adaptation.

104