The inference of sex-biased human demography from whole-genome data

• DOI: 10.1371/journal.pgen.1008293
• Publication Date: 2019-09-20T17:37:28Z
• Authors: Shaila Musharoff; Suyash Shringarpure; Carlos D. Bustamante; Sohini Ramachandran
• Keywords: Effective population size; Demographic history; Sampling bias; 1000 Genomes Project

Sex-biased demographic events ("sex-bias") involve unequal numbers of females and males. These are typically inferred from the relative amount X-chromosomal to autosomal genetic variation have led conflicting conclusions about human history. Though population size changes alter diversity even in absence sex-bias, this has generally not been accounted for sex-bias estimators date. Here, we present a novel method identify sequence data that models estimates female fraction effective during each time epoch. Compared recent inference methods, our approach can detect on single branch without requiring an outgroup or knowledge divergence events. When applied simulated data, conventional biased by changes, especially growth bottlenecks, while estimator is unbiased. We next apply high-coverage exome 1000 Genomes Project estimate male bias Yorubans (47% female) Europeans (44%), possibly due stronger background selection X chromosome than autosomes. Finally, Phase 3 Complete Genomics whole-genome (63% female), (84%), Punjabis (82%), as well Peruvians (56%), Southern Han Chinese (45%). Our additionally identifies male-biased migration out Africa based (20% female). results demonstrate modeling change necessary parameters accurately. gives insight into signatures sexual species, it produces serve more accurate null tests selection.