REVIEWS Recent and ongoing selection in the human genome Rasmus Nielsen*, Ines Hellmann*, Melissa Hubisz‡, Carlos Bustamante§ and Andrew G. Clark|| Abstract | The recent availability of genome-scale genotyping data has led to the identification of regions of the human genome that seem to have been targeted by selection. These findings have increased our understanding of the evolutionary forces that affect the human genome, have augmented our knowledge of gene function and promise to increase our understanding of the genetic basis of disease. However, inferences of selection are challenged by several confounding factors, especially the complex demographic history of human populations, and concordance between studies is variable. Although such studies will always be associated with some uncertainty, steps can be taken to minimize the effects of confounding factors and improve our interpretation of their findings. Genetic drift The past few years have seen an explosion of studies pattern. Therefore, it might be possible to identify putative The stochastic change in using molecular data to detect Darwinian natural genetic disease factors by identifying regions of the human population frequency of a selection1–6. With the recent availability of large-scale genome that currently are under selection3,11. In general, mutation due to the sampling genotyping data, genome-wide scans for genes or positions in the genome that are under selection must process that is inherent in reproduction. genomic regions that have been targeted by selec- be of functional importance, otherwise selection could tion have become feasible. These studies have greatly not be operating. Adaptation advanced our understanding of human evolution and The aim of this Review is to discuss some of the major Heritable changes in genotype molecular evolution in general, but they have also findings regarding selection in humans, and explain or phenotype that result in sparked considerable controversy. why the conclusions of these studies have at times been increased fitness. The interest in detecting selection is twofold. First, it controversial with low levels of concordance among stems from a natural curiosity about our evolutionary past studies. We focus particularly on recent selection; that and the basic mechanisms that govern molecular evolu- is, selection that might have affected current population tion. Much of the work in this field through the past four genetic variation. We first address the question of the *Center for Comparative decades has focused on quantifying the relative impor- likely relative contributions of negative and positive Genomics, University of genetic drift Copenhagen, tance of Darwinian selection and random selection to genetic variation in human populations, and Universitetsparken 15, in determining levels of variability within species, explore how identifying these types of selection might 2100 Kbh Ø, Denmark. as well as divergence between species (for example, contribute to our understanding of human evolution and ‡Department of Human REFS 7,8). However, as evidence accumulates for a strong gene function. We then discuss the different approaches Genetics, University of role of selection, efforts are increasingly concentrating that are taken to detecting the recent and ongoing Chicago 920 E. 58th Street, Chicago, Illinois 60637, USA. on identifying and characterizing particular instances of positive selection that has affected the human genome, §Department of Biological selection and adaptation at the molecular level. In humans, followed by a detailed discussion of recent genome-wide Statistics and Computational in particular, there has been a strong interest in identify- studies that have provided many new potentially selected Biology, Cornell University, ing genes that have undergone recent selection relating to genomic regions and individual genes. The key problems 1198 Comstock Hall, Ithaca, 4,9,10 New York 14853, USA. key human traits such as cognitive abilities . that face studies of selection are then addressed, along ||Department of Molecular A second motivation for studying selection stems with a discussion of why low concordance has been seen Biology and Genetics, from the realization that inferences about selection among some of the studies that have been carried out Cornell University, can provide important functional information. For so far. Finally, we bring together a discussion of studies 107 Biotechnology Building, example, genes that are targeted by selection acting on that have provided insights into the patterns of selection Ithaca, New York 14853, USA. Correspondence to R.N. segregating mutations are more likely to be associated that are likely to characterize Mendelian and complex e-mail: [email protected] with disease (for example, REF. 3). Even small fitness human diseases, with the potential to aid the discovery doi:10.1038/nrg2187 effects can, on an evolutionary timescale, leave a distinct of further disease-associated genes. NatUre reViewS | GENETICS VOLUme 8 | NOVemBer 2007 | 857 © 2007 Nature Publishing Group REVIEWS Positive and negative selection have recently taken advantage of this fact to quantify Although it is clear that selection is pervasive in humans positive selection acting in the genome on the human and other organisms, the relative importance of posi- lineage leading from the ancestor of human and chim- tive and negative selection is still debated. Much of the panzees to modern humans18–20. In general, these stud- natural selection acting on genomes may be negative ies have identified genes involved in immune-related selection acting to remove new deleterious mutations. functions, spermatogenesis, olfaction and sensory per- Most exons in protein-coding regions are highly con- ception, and have highlighted several other functional served between species, because many potential muta- gene categories with an increased likelihood of having tions would disrupt protein function. Therefore, the experienced positive selection. Genes in these categories conservation of genic regions provides evidence of past are likely to be involved in direct interactions with the negative selection and provides an important route to environment, and will be under selective pressure in genome annotation. Similar evidence for conservation the face of environmental change. In particular, genes and negative selection in non-coding regions provides involved in dynamic competitive or co-evolutionary the basis for an important approach for detecting interactions are expected to experience more positive functional elements, such as microRNAs (for example, selection. A prime example of this is immunity and REFS 12,13). defence-related genes, which are involved in dynamic Eyre-Walker and Keightley14 estimated that at least interactions with pathogens. As a category, these genes 38% of all new amino-acid altering mutations in the have experienced by far the most positive selection in human genome are being eliminated by negative selec- humans and other organisms18–20. tion, assuming that all mutations are either deleteri- There are several theories regarding selection act- ous or neutral (that is, having no effect on organismal ing on spermatogenesis, one being that most selection fitness). As noted by the authors of this study, this is is related to post-mating competition between sperm probably an underestimate, and subsequent studies15–17 from different males for fertilization21. In this case, the have suggested that as a much as 70–75% of amino-acid changing environment is the phenotype of sperm from altering mutations are affected by moderate or strong other males. Alternative theories suggest that the selec- negative selection. Importantly, however, much of this tion is related to interactions between egg and sperm selection might act at the level of gametogenesis, on cells22,23, or that it is driven by selfish mutations causing mature gametes or during early development. Mutations segregating distortion20. that are strongly deleterious will be quickly eliminated Several individual genes that might underlie human- by natural selection, and only mutations that have, at specific adaptations have been highlighted in interspecies worst, a mildly negative fitness effect will be observed studies, including genes involved in speech and cogni- Fitness as segregating in the population. A. R. Boyko et al. tion, such as forkhead box P2 (FOXP2), genes associ- A measure of the capacity of (unpublished observations) estimated that the propor- ated with pregnancy, such as the progesterone receptor an organism to survive and tion of amino-acid altering mutations in humans that (PGR), genes associated with skeletal development, such reproduce. have a negative fitness effect, but are so weakly selected as tolloid-like 2 (TLL2), and numerous other genes4,18. that they might still be segregating in the population, is However, genomic comparisons between species alone Effective population size (Ne) The size of a population approximately 30–40%. do not inform us about ongoing and recent selection measured by the expected Positive selection occurs when a new (or previ- within species, and have little power if genes have been effect (through genetic drift) of ously rare) mutation confers a fitness advantage to the affected by only a single, recent selective event, even the population size on genetic individuals carrying it. Much attention has focused on if the strength of selection acting on the mutation is variablity. Ne is typically much