北京雁栖湖应用数学研究院

韩立岩 , 李振 , 刘庆富 , 龙飞 , 汤珂

Scott Edwards

2025年01月17日 16:00 至 18:00

A7-302

Zoom 230 432 7880 (BIMSA)

Pangenomes are collections of genomes that have been assembled and constructed in ways that faithfully capture all types of genomic variation, including single-nucleotide polymorphisms but also individual haplotypes and structural variation (SV), such as copy number variants, insertions and deletions, and complex variation in multigene families like the major histocompatibility complex (MHC). In this talk I will present an analysis of 45 genomes from three species of Scrub-Jay (Aphelocoma), each assembled with PacBio HiFi reads and the long-read genome assembler hifiasm. Heterozygosity and PSMC analysis show that the three species fall along a 60-fold gradient of effective population sizes that allow us to observe the differing dynamics of SVs and SNPs in the outbred Woodhouse’s Scrub-Jay, lower diversity Florida Scrub-Jay and the inbred Island Scrub Jay. These genomes are relatively large for birds (~1.3 Gb) and harbor an unusually high proportion of transposable elements (up to ~27%). We find compelling evidence for a rapid reduction of genome size by ~80 Mb) in the Island Scrub-Jay (A. insularis) due primarily to reductions of major satellite families. Using pangenome graphs, we found that indels (< 50 bp; ~3.1M) and SVs (> 50 bp; ~448,000), including 382 inversions, were drastically less common than SNPs (~20.3M) across the species complex and reveal that SV length correlates with the severity of estimated SV fitness effects. Telomeres can be measured fairly accurately and are shorter in Island Scrub-Jays, consistent with population genetic predictions for small populations. In Woodhouse’s Scrub Jay, we demonstrate effects of SVs on gene expression, due both to direct ablation of genes and via removal of putative noncoding enhancers. Pangenomes offer an exciting new way to monitor genetic variation and its fitness consequences in natural populations of birds.