Fig. 1

Architecture of polymorphisms in the human genome. a Percentage of polymorphisms in different regions (5′UTRs, coding regions, introns, 3′UTRs) of the human genes. b Average SNV and INDEL densities (# polymorphisms/kb) in the different regions (5′UTR, coding region, intron and 3′UTR) of a gene in the human genome. Error bars represent the standard errors of the mean SNV and INDEL densities. c Percentage of frame-shift and in-frame INDELs in coding and non-coding regions in human genes. Frame-shift INDELS are defined as INDELs whose lengths are not in multiples of three while in-frame INDELs are those whose lengths are in multiples of three. d Percentage of SNVs with different DAFs in the four genic regions, as measured in HapMap individuals from African, Asian and European populations. e Percentage of synonymous and non-synonymous variants with different DAFs, as measured in HapMap individuals from African, Asian and European population groups. f Distribution of F_ST statistics in four genic regions. SNVs in coding regions show significantly lower median F_ST compared to the other non-coding regions. g Fold enrichment of SNVs showing signatures of negative selection (F_ST = 0) (open bar) or RPS (shaded bar) in the genic regions. Fold enrichment is determined by the percentage of SNVs with F_ST = 0 or under RPS in a specific region (e.g. coding region) divided by the percentage of all genotyped SNVs in that region. Coding, coding region; non-syn, non-synonymous; syn, synonymous. AFR, African; ASN, Asian; EUR, European. ***p < 0.001, **p < 0.01, *p < 0.05; ns, not significant