The difficulty of avoiding false positives in genome scans for natural selection

Abstract

Several studies have found evidence for more positive selection on the chimpanzee lineage compared with the human lineage since the two species split. A potential concern, however, is that these findings may simply reflect artifacts of the data: inaccuracies in the underlying chimpanzee genome sequence, which is of lower quality than human. To test this hypothesis, we generated de novo genome assemblies of chimpanzee and macaque and aligned them with human. We also implemented a novel bioinformatic procedure for producing alignments of closely related species that uses synteny information to remove misassembled and misaligned regions, and sequence quality scores to remove nucleotides that are less reliable. We applied this procedure to re-examine 59 genes recently identified as candidates for positive selection in chimpanzees. The great majority of these signals disappear after application of our new bioinformatic procedure. We also carried out laboratory-based resequencing of 10 of the regions in multiple chimpanzees and humans, and found that our alignments were correct wherever there was a conflict with the published results. These findings throw into question previous findings that there has been more positive selection in chimpanzees than in humans since the two species diverged. Our study also highlights the challenges of searching the extreme tails of distributions for signals of natural selection. Inaccuracies in the genome sequence at even a tiny fraction of genes can produce false-positive signals, which make it difficult to identify loci that have genuinely been targets of selection.

Overestimation of nonsynonymous/synonymous rate ratio by reverse-translation of aligned amino acid sequences

“…These results suggest that even if the proportion of misaligned codon sites is small, they cause significant overestimation of the d N /d S ratio for the entire alignment of codon sequences constructed by reverse-translating aligned amino acid sequences (Wong et al, 2008;Mallick et al, 2009;Schneider et al, 2009). These codon sites may also be falsely identified as positively selected (Vamathevan et al, 2008;Wong et al, 2008;Mallick et al, 2009;Schneider et al, 2009;Fletcher and Yang, 2010).…”

“…However, the d N /d S ratio is known to be inflated at mis-aligned codon sites (Wong et al, 2008;Mallick et al, 2009;Schneider et al, 2009). It has also been reported that as more non-homologous codons are aligned, more amino acid sites are falsely identified as positively selected (Vamathevan et al, 2008;Wong et al, 2008;Mallick et al, 2009;Schneider et al, 2009;Fletcher and Yang, 2010). The purpose of the present study was to examine the effect of reverse-translating aligned amino acid sequences on the estimation of d N /d S ratio, through a large-scale analysis of protein-coding nucleotide sequences from vertebrate species.…”

Statistical methods for detecting natural selection from genomic data

“…The existence of a single codon site with the occurrence of multiple nonsynonymous substitutions in the entire sequence sometimes appeared to be sufficient for erroneously detecting positive selection. This method was also found to produce spurious results of positive selection when estimates of d N /d S were inflated due to episodes of GC-biased gene conversion (Berglund et al, 2009;Galtier et al, 2009;Hurst, 2009) and each of sequencing, annotation, and alignment errors (Mallick et al, 2009;Schneider et al, 2009;Fletcher and Yang, 2010). From the analysis of rhodopsins in vertebrates, it was shown that positively selected sites identified by the branch-site method mostly disagreed with experimentally determined adaptive sites (Nozawa et al, 2009b).…”

“…Kelley et al (2006) also conducted computer simulation based on Tajima's D and found that the false-discovery rate can be high when positive selection is weak, the fraction of truly positively selected loci is small, and the fraction of test statistic values considered as evidence for positive selection is large. Little overlap among results from different studies It has been reported that the fraction of overlap between positively selected regions identified in different studies is generally small (Mallick et al, 2009). When 713 positively selected genes in human identified by the iHS test (Voight et al, 2006) were compared with 90 positively selected genes identified by the LDD test (Wang et al, 2006), only 9 genes were shared between the two results .…”

“…Reverse-tr an slation of align ed am in o acid sequences In the analysis of natural selection by comparing d S and d N , homologous codons are often identified by reverse-translating aligned amino acid sequences into codon sequences. In this method, however, homologous codons may be mis-identified when frame-shifts occurred (Mills et al, 2006) or amino acid sequences were misaligned (Wong et al, 2008;Liu et al, 2009;Fletcher and Yang, 2010), which may lead to overestimation of the d N /d S ratio (Wong et al, 2008;Mallick et al, 2009;Schneider et al, 2009) and false detection of positively selected amino acid sites (Vamathevan et al, 2008;Wong et al, 2008;Mallick et al, 2009;Schneider et al, 2009;Fletcher and Yang, 2010). In a large-scale analysis of protein-coding nucleotide sequences from vertebrate species, it was observed that 1-9% of codon sites that were identified as homologous with reverse-translation apparently contained non-homologous codons, where the d N /d S ratio was unduly high, and the overall d N /d S ratio was 5-43% overestimated with reverse-translation (Y.…”