November 2026 talk of the monthly meeting of the #animal-genomics special interest group.

We are happy to introduce Ekaterina Noskova (Institute of Ecology and Evolution, University of Edinburgh, UK). Ekaterina Noskova is an MSCA Postdoctoral Fellow at the Institute of Ecology and Evolution, University of Edinburgh. She works on statistical and computational methods for inferring the evolutionary history of populations from genetic data. During her PhD she developed GADMA, a widely used tool for automatic demographic inference. Before Edinburgh, she held postdoctoral positions at ETH Zurich and the University of Fribourg in Switzerland.

Genome-wide joint inference of demography and linked selection from time-series data

Time-series allele frequency data, particularly ancient DNA, are among the richest sources of information about recent selection. Yet most existing methods treat loci independently, inferring site-specific selection coefficients in isolation and relying on ad hoc window statistics to account for hitch-hiking. Methods that explicitly model linkage remain computationally demanding at genome scale. I will present SweepLink, which jointly infers demography and linked selection across the genome by allowing neighbouring loci to share information about selection. This pooling suppresses drift-driven false signals and improves sensitivity to weak and moderate selection. Applied to ancient human DNA from Britain, SweepLink recovers the canonical targets of recent European adaptation and identifies candidate regions missed by a previous single-locus scan of the same data.