The promise and limitations of population exomics for human evolution studies

Abstract

Exome sequencing is poised to yield substantial insights into human genetic variation and evolutionary history, but there are significant challenges to overcome before this becomes a reality.

Figure 1

The contributions of different data types to population genetic inferences. Exomes, SNP arrays, and genomes are likely to capture different combinations of these four basic site types: common variants, rare variants, invariant sites, and noncoding sites. Each site type offers unique information relevant to analyses of demography and natural selection. Rare variants and invariant sites captured in exomes are important for numerous evolutionary questions, from estimating the effective population size to detecting positive selective sweeps, and they may be missed by other methods. However, phenotypically causal noncoding variants and truly neutral regions far from genes and free from the effects of selection are absent from exomes.

Figure 2

The influence of sample size in detecting recent population expansions. (a) Demographic model used in the simulations. Times of demographic events (in units of thousands of years, kyr) and population sizes are indicated. (b) The site frequency spectrum for sample sizes of 50 and 2,000 chromosomes. (c) Average value of Tajima's D based on 10⁴ simulation replicates as a function of sample size (number of chromosomes). The lines denote values of Tajima's D from the demographic model in (a) (black line) and a constant sized population (green line). This simulation illustrates the power of large sample sizes for inferring recent demographic events. All information in (b, c) is based on simulating 10 kb of sequence.