The effect of SNP discovery method and sample size on estimation of population genetic data for Chinese and Indian rhesus macaques (Macaca mulatta)

Abstract

This study was designed to address issues regarding sample size and marker location that have arisen from the discovery of SNPs in the genomes of poorly characterized primate species and the application of these markers to the study of primate population genetics. We predict the effect of discovery sample size on the probability of discovering both rare and common SNPs and then compare this prediction with the proportion of common and rare SNPs discovered when different numbers of individuals are sequenced. Second, we examine the effect of genomic region on estimates of common population genetic data, comparing markers from both coding and non-coding regions of the rhesus macaque genome and the population genetic data calculated from these markers, to measure the degree and direction of bias introduced by SNPs located in coding versus non-coding regions of the genome. We found that both discovery sample size and genomic region surveyed affect SNP marker attributes and population genetic estimates, even when these are calculated from an expanded data set containing more individuals than the original discovery data set. Although none of the SNP detection methods or genomic regions tested in this study was completely uninformative, these results show that each has a different kind of genetic variation that is suitable for different purposes, and each introduces specific types of bias. Given that each SNP marker has an individual evolutionary history, we calculated that the most complete and unbiased representation of the genetic diversity present in the individual can be obtained by incorporating at least 10 individuals into the discovery sample set, to ensure the discovery of both common and rare polymorphisms.

4040 SNPs for genomic analysis in the rhesus macaque (Macaca mulatta)

“…Although the differentiation between the Indian and Chinese samples was quite strong and consistent with previously reported results [4, 23], two results unique to the present study are the greater heterozygosity in Indian than in Chinese rhesus macaques and the presence of population structure of unknown source in the Indian sample (Figures 4 and 5). Previous studies of microsatellite (STR) markers [43], mtDNA sequence [22] and SNPs in the 3’ ends of rhesus macaque genes [3] and ENCODE regions [33] have reported higher levels of heterozygosity in Chinese than in Indian rhesus macaques.…”

“…A much greater proportion of the markers examined in our study were polymorphic in both the Indian and Chinese samples, and the proportion of markers that were population specific much smaller. This difference is probably largely due to a bias towards the discovery of high MAF, shared SNPs inherent in our discovery method [23]: the comparison of just two individuals, one Indian and one Chinese, led to preferential discovery of high frequency polymorphisms shared between the two populations. Low frequency polymorphisms had a lower probability of discovery, while discovery of SNPs polymorphic only in the Indian or Chinese populations required the sequenced individual to be a heterozygote.…”

“…These animals were either imported directly from the country of origin (VBS, TSS) or had sufficient colony documentation to support their assignment to the appropriate region (CPRC, UM, ONPRC and CNPRC). In addition, all individuals had been sequenced for a 830-bp section of the mitochondrial genome [21] and 24 nuclear microsatellite loci [22], used to confirm their assignment to a specific geographic region (partial data presented in Satkoski Trask et al [23], additional data not shown). Approximately 5% of the markers and at least two individuals were chosen at random and duplicated across each OPA and each run as controls.…”

Patterns of genetic variation and the role of selection in HTR1A and HTR1B in macaques (Macaca)

“…These regions have no known function and are thus referred to in this paper as the “nonfunctional regions,” so as to differentiate these regions from the noncoding regions surrounding the serotonin genes, the latter of which may have potential regulatory roles. Because these nonfunctional regions are at least 20,000 base pairs from the nearest coding region [80], they are presumed to be evolving neutrally. The PCR protocols differed for each of the regions amplified and are available upon request.…”

Variation in the chickenLPIN2gene and association with performance traits

“…Based on Trask's suggested sample size for ensuring the discovery of both common and rare polymorphisms in a population (Trask et al, 2011), 10 individuals were randomly selected (as unrelated as possible) from each breed. Anticoagulant (heparin sodium) blood was collected from a wing vein and kept at −20°C.…”