A comprehensive detection and characterization of deleterious variants in traditional breeds

Chiara Bortoluzzi1*, Martijn Derks1, Steffen Weigend2, Martien AM Groenen1, Hendrik-Jan Megens1

1 Animal Breeding and Genomics Centre, Wageningen University and Research, The Netherlands 2 Friedrich Loeffler Institute (FLI), Department of Breeding and Genetical Resources, Greifswald, Germany

Background Materials & Methods Deleterious variants are constantly generated by mutation (Fig 1), but are not always purged from the population. Whole genome sequence Mapping & population data (WGS) variant calling Understanding the extent and the nature of deleterious variation in small populations is of interest for conservation purposes, because Phred quality score/ depth of coverage Variant filter these variants reduce reproductive fitness (Fig 2) and genetic Ancestral state of variant diversity when homozygous. 60K SNP chip (78 False discovery rate samples)

Synonymous / Missense Variant effect predictor tolerated (VEP) fail to hatch

Number of eggs that Frameshift, splice variant, Inbreeding Inbreeding coefficient (F) stop gained, missense Deleterious variants deleterious, inframe Figure 1. An example of deleterious variants Figure 2. Reduction in reproductive fitness due to the accumulation of Figure 3. Pipeline overview used to detect deleterious variants using whole-genome sequence data deleterious variants as result of inbreeding

Traditional chicken breeds offer a powerful model to address the role of Discussion demographic discontinuities and selection on deleterious variants The predicted deleterious variants are subject to purifying selection, as confirmed by the skewed distribution towards a higher proportion of low- Results high derived allele frequency alleles (Fig. 4A, Fig. 4B). Frameshift 0.6 Splice Inframe Start−stop Deleterious Mutational load considerably varies across traditional breeds as a result of 0.5

Tolerated 0.5 Synonymous their different demographic and management history (Fig. 5). 0.4 0.4 Genotypes containing putatively derived alleles are lowest for deleterious and 0.3

0.3 LoF variants (Fig. 6) relative to synonymous and missense tolerated (Fig. 7). Relative fraction Relative

0.2 This further suggests that purifying selection effectively removes deleterious 0.2 Relative fraction Relative variation even in population characterised by small effective population size. 0.1

0.1

0.0 Synonymous" Nonsynonymous"tolerated" 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 55000"

0.0 Derived allele frequency 50000" 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 45000" 40000" Derived allele frequency Figure 4B. Derived allele frequency of loss 35000" of function (LoF) variants 30000" Figure 4A. Derived allele frequency of 25000" 20000" synonymous and deleterious variants 1400" Nonsynonymous"deleterious" Loss"of"function" 15000" 10000"

1200" Total&number&of&derived&alleles& 5000" 1000" 0"

800" 0.012 " " Breda"Fowl" Chaam"Fowl" Dutch"Fowl" " Schijndelaar" " Drenthe"Fowl" Frisian"Fowl"" Kraienkoppe" 600" Assendelft"Fowl" Groninger"Mew" Dutch"Owl"Bearded" Barnevelder"Bantam"Breda"Fowl"Bantam" Dutch"Fowl"Bantam" Eikenburger"Bantam"Frisian"Fowl"Bantam" Lakenvelder"Bantam"Schijndelaar"Bantam"Welsummer"Bantam" Drenthe"Fowl"Bantam"Dutch"Booted"Bantam" Dutch"Polish"Bearded" Assendelft"Fowl"Bantam" Groninger"Mew"Bantam" 400" Dutch"Polish"non"Bearded"Dutch"Polish"non"Bearded" 0.010 Dutch"Owl"Bearded"Bantam" North"Holland"Blue"Bantam" 200" Total&number&of&derived&alleles& Figure 7. Total number of derived (heterozygous and homozygous 0" 0.008 derived) alleles for synonymous and missense tolerated variants

Brabanter" Breda"Fowl" Dutch"Fowl" Barnevelder" Chaam"Fowl" Frisian"Fowl"" Kraienkoppe"Lakenvelder" Schijndelaar" Welsummer" 0.006 ● Drenthe"Fowl" Assendelft"Fowl" Groninger"Mew"

Mutational load (nonsynonymous/synonymous ratio) Mutational load (nonsynonymous/synonymous Breda"Fowl"Bantam" Dutch"Fowl"Bantam"Dutch"Owl"Bearded" Barnevelder"Bantam" Eikenburger"Bantam"Frisian"Fowl"Bantam" Lakenvelder"Bantam"Schijndelaar"Bantam"Welsummer"Bantam" Drenthe"Fowl"Bantam"Dutch"Booted"Bantam" Dutch"Polish"Bearded" Assendelft"Fowl"Bantam" Groninger"Mew"Bantam" Dutch"Polish"non"Bearded" Dutch"Owl"Bearded"Bantam" North"Holland"Blue"Bantam" Dutch"Polish"non"Bearded"Bantam"

Brabanter Figure 6. Total number of derived (heterozygous and homozygous Welsummer Barnevelder Breda_Fowl Lakenvelder Dutch_Fowl Schijndelaar Kraienkoppe Frisian_Fowl Chaam_Fowl Drenthe_Fowl Groninger_Mew Assendelft_Fowl derived) alleles for deleterious and loss of function variants Welsummer_Bantam Barnevelder_Bantam Dutch_Owl_Bearded Eikenburger_Bantam Breda_Fowl_Bantam Lakenvelder_Bantam Dutch_Fowl_Bantam Schijndelaar_Bantam Frisian_Fowl_Bantam Dutch_Polish_Bearded Drenthe_Fowl_Bantam Dutch_Booted_Bantam Groninger_Mew_Bantam Assendelft_Fowl_Bantam

Dutch_Polish_non_Bearded North_Holland_Blue_Bantam Dutch_Owl_Bearded_Bantam

Dutch_Polish_non_Bearded_Bantam We developed a comprehensive catalogue of putatively deleterious variants for traditional Dutch chicken Figure 5. Mutational load of homozygous derived breeds. Such genomic catalogues can be used to enhance the conservation status of traditional breeds by lowering alleles the frequency of undesirable variant in the population.

* Contact author: [email protected] | Animal Breeding and Genomics Centre, P.O. Box 123, 6708PB Wageningen