Comprehensive analysis of Japanese population history by detecting ancestry-marker polymorphisms without using ancestral genomic information

Modern Japanese have two major ancestral populations: the indigenous Jomon hunter gatherers and continental East Asian farmers. To figure out the formation process of current Japanese population, we developed a reference-free detection method of variants derived from ancestral populations using a summary statistic, the ancestry-marker index (AMI). We confirmed by computer simulations that AMI can detect ancestry-derived variants even in an admixed population of recently diverged source populations with high accuracy, which cannot be achieved by the most widely used statistics, S*, for identifying archaic ancestry. We applied the AMI to modern Japanese samples and identified 208,648 single nucleotide polymorphisms (SNPs) that were likely derived from the Jomon people (Jomon-derived variants). The analysis of Jomon-derived variants in 10,842 modern Japanese individuals recruited from all over Japan revealed that the admixture proportions of the Jomon people varied between prefectures, probably due to the differences of population sizes of immigrants in the final Jomon to the Yayoi period. The estimated allele frequencies of genome-wide SNPs in the ancestral populations of modern Japanese suggested their phenotypic characteristics possibly for adaptation to their respective livelihoods; higher triglycerides and blood sugar for the Jomon ancestry and higher C-reactive protein and eosinophil counts for continental ancestry. According to our findings, we propose a formation model of modern Japanese population; regional variations in admixture proportions of the Jomon people and continental East Asians formed genotypic and phenotypic gradations of current Japanese archipelago populations.