SNP-based Computational Analysis Reveals Recombination-associated Genome Evolution in Humans

Abstract

Background: Meiotic recombination is an important source of genetic variation, but how recombination shapes the genome is not clearly understood yet. Objective: Here, we investigate the roles of recombination on human genome evolution from two aspects: How does recombination shape single nucleotide polymorphism (SNP)-related genomic variation features? Whether recombination drives genome evolution through a neighbor-dependent mutational bias? Methods: We analyzed the relationship of recombination rate with mutational bias and selection effect at SNP sites derived from the 1000 Genomes Project. Results: Our results show that SNP density, Ts/Tv, nucleotide diversity, and Tajima's D were positively correlated with the recombination rate, while Ka/Ks were negatively correlated with the recombination rate. Moreover, compared with non-coding regions, gene exonic regions have lower nucleotide diversity but higher Tajima's D, suggesting that coding regions are subject to stronger negative selection but have fewer rare alleles. Gene set enrichment analysis of the protein-coding genes with extreme Ka/Ks ratio implies that under the effect of high recombination rates, the genes involved in the cell cycle, RNA processing, and oocyte meiosis are subject to strong negative selection. Our data also support S (G or C) > W (A or T) mutational bias and W>S fixation bias in high recombination regions. In addition, the neighbor-dependent mutational bias was found to be stronger at high recombination regions. Conclusion: Our data suggest that genetic variation patterns, particularly the neighbor-dependent mutational bias at SNP sites in the human genome, are mediated by recombination.