Preprint delivers faster, more accurate method for mapping how geography shapes genetic relatedness
A new computational approach called Deep Recovery of Effective Migration Surfaces reduces the computational cost of fitting isolation-by-distance models to spatially referenced genotype data.
A preprint posted to bioRxiv presents a faster method for fitting effective migration surface (EMS) models to spatially referenced genomic data. EMS models visualise how the strength of isolation-by-distance — the tendency for genetic relatedness to decline with geographic distance — varies across a landscape, potentially revealing barriers such as mountain ranges, rivers, or historical population bottlenecks.
Fitting likelihood-based EMS models has historically been computationally demanding, limiting their application to large datasets or fine-scale geographic analyses. The authors, who describe their approach as 'Deep Recovery', report improvements in both computational efficiency and accuracy of surface estimation, benchmarked against existing methods.
Applications of EMS methods span conservation genetics, human population history, and the study of non-human species. Faster fitting opens the door to larger sample sizes and more granular geographic grids, potentially improving the resolution with which migration barriers and corridors can be detected from genomic data.
This preprint has not been peer-reviewed.
Sources
Read the original reporting — these are the public sources this summary draws from.
-
Primary sourcePreprint bioRxiv · 2026-11-04Deep Recovery of Effective Migration Surfaces