Toward male individualization with rapidly mutating y-chromosomal short tandem repeats

Abstract

Relevant for various areas of human genetics, Y-chromosomal short tandem repeats (Y-STRs) are commonly used for testing close paternal relationships among individuals and populations, and for male lineage identification. However, even the widely used 17-loci Yfiler set cannot resolve individuals and populations completely. Here, 52 centers generated quality-controlled data of 13 rapidly mutating (RM) Y-STRs in 14,644 related and unrelated males from 111 worldwide populations. Strikingly, >99% of the 12,272 unrelated males were completely individualized. Haplotype diversity was extremely high (global: 0.9999985, regional: 0.99836-0.9999988). Haplotype sharing between populations was almost absent except for six (0.05%) of the 12,156 haplotypes. Haplotype sharing within populations was generally rare (0.8% nonunique haplotypes), significantly lower in urban (0.9%) than rural (2.1%) and highest in endogamous groups (14.3%). Analysis of molecular variance revealed 99.98% of variation within populations, 0.018% among populations within groups, and 0.002% among groups. Of the 2,372 newly and 156 previously typed male relative pairs, 29% were differentiated including 27% of the 2,378 father-son pairs. Relative to Yfiler, haplotype diversity was increased in 86% of the populations tested and overall male relative differentiation was raised by 23.5%. Our study demonstrates the value of RM Y-STRs in identifying and separating unrelated and related males and provides a reference database.

Keywords: RM Y-STRs; Y-STRs; Y-chromosome; forensic; haplotypes; paternal lineage.

Figure 1

Two-dimensional plot of MDS analysis of Slatkin's linearized F_ST values for RM Y-STR haplotypes in a global sample of 12,272 individuals from 111 populations (stress 0.07462). Smaller inset MDS shows the effect of equalized sample size (N = 100 individuals or fewer per population, stress = 0.02416).

Figure 2

Proportion of individuals with haplotypes shared within populations (right) and between populations (left) for Yfiler (light blue bars) and the RM Y-STR set (dark red bars).

Figure 3

Geographic representation of pairwise between-population haplotype sharing. Blue lines connect population pairs showing shared haplotypes for A: Yfiler, and B: the RM Y-STRs set. Smaller insets show Europe enlarged.

Figure 4

Weighted median-joining networks from single-copy Y-STRs for 1,000 individuals randomly selected from the total dataset, with regional geographic groups indicated by colors (see inset legend). A: Based on 15-loci Yfiler haplotypes (excluding DYS385a/b). B: Based on 10-loci RM Y-STR haplotypes (excluding DYF399S1, DYF403S1, and DYF404S1).

Figure 5

Empirical male relative differentiation using 2,528 paternal relative pairs separated by one to 20 meioses for Yfiler (light blue bars) and the RM Y-STR set (dark red bars). The data combine the new 2,372 relative pairs from the current study with the previously used 156 relative pairs [Ballantyne et al., 2012]. Values above the bars indicate the absolute number of relative pairs the estimated differentiation rate is based upon. Error bars represent binomial confidence intervals.