New Method Accurately Estimates Demography and Mutation Rates from Million-Haploid Genomes

A new study published in The American Journal of Human Genetics introduces a novel computational method, dubbed DR EVIL, designed to more accurately estimate demographic history and mutation rates from ultra-large genomic datasets. The research, led by Joshua G. Schraiber, Jeffrey P. Spence, and Michael D. Edge, addresses critical limitations of traditional population genetics tools when applied to modern sequencing data, particularly concerning rare genetic variants.

The paper, titled "Estimation of demography and mutation rates from one million haploid genomes," highlights that current methods often fail to account for recurrent mutations, where the same genetic variant arises multiple times independently. This oversight can lead to biased estimates of population history and mutation rates, especially in datasets comprising millions of genomes. DR EVIL overcomes this by employing a diffusion approximation to a branching-process model, allowing for the inclusion of recurrent mutations and natural selection in its calculations.

A key finding from the application of DR EVIL to one million haploid samples from the gnomAD v4.1 dataset is the refined estimation of recent demographic history, indicating an effective population size of approximately 25 million individuals over the last 10 generations for the non-Finnish European cohort. This estimate aligns with other recent findings and underscores the method's accuracy. The study also identified significant mutation-rate heterogeneity across different genomic contexts, even after accounting for factors like trinucleotide context and methylation status.

The authors emphasize that at current sample sizes, many polymorphic sites with high mutation rates are likely to represent descendants of multiple mutation events, a phenomenon often overlooked by conventional approaches. This new methodology provides a more robust framework for understanding human evolutionary history and the dynamics of genetic variation, offering improved accuracy in estimating fundamental population genetics parameters. The American Journal of Human Genetics, a peer-reviewed publication of the American Society of Human Genetics, is recognized for its significant contributions to the field.