Austin Daigle
banner
adaigle.bsky.social
Austin Daigle
@adaigle.bsky.social
61 followers 100 following 6 posts
PhD Candidate in Bioinformatics & Computational Biology at UNC, coadvised by Dan Schrider & Parul Johri. I work on population genetics, transposon detection, and simulation-based inference—simulating evolution because real evolution takes way too long.
Posts Media Videos Starter Packs
Reposted by Austin Daigle
Reposted by Austin Daigle
Amid the chaos, it was great to share results from the first Genomic History Inference Strategies Tournament at ProbGen25. Watch the talk to learn about the future of GHIST and @andrewhvaughn.bsky.social's nefarious metagaming of demographic inference. 😀 www.youtube.com/watch?v=3_dv...
Talk describing the results of GHIST1 at ProbGen 25.
YouTube video by Ryan Gutenkunst
www.youtube.com
These results highlight a key challenge for population-genetic analyses in highly selfing species or low-recombination genomic regions. Check out the paper for a deeper dive into other potentially relevant factors like beneficial mutations, dominance coefficients, and population structure!
If HRI was the cause mis-inference, we’d expect it to be caused by the reduced levels of effective recombination in selfers. Indeed, when we ran simulations with lower recombination (but no selfing), we saw the same patterns of DFE mis-inference.
We hypothesized this mis-inference was caused by HRI, where linked deleterious mutations interact and reduce the efficacy of selection. The site frequency spectrum (SFS) had a U-shape at high selfing rates, a pattern often linked to HRI and not modeled by current DFE inference approaches.
In simulated highly selfing populations, the DFE was mis-inferred by two unique DFE inference methods—nearly neutral and strongly deleterious mutations were overestimated, while mildly deleterious ones were underestimated.
Excited to share my first PhD project with my mentor, @johriparul.bsky.social! We examine how Hill–Robertson interference (HRI) in highly selfing species biases estimates of the distribution of fitness effects of new mutations (DFE).
doi.org/10.1093/evol...
@journal-evo.bsky.social #popgen #evobio
Thrilled to have been part of the inaugural GHIST competition in population genetics inference! Big thanks to the organizers for a fun and challenging event. Congrats to Andrew Vaughn and Ekaterina Noskova for their impressive performance, I'm looking forward to reading about everyone's methods.
Congrats to Andrew Vaughn @andrewhvaughn.bsky.social, for winning 3 out of 4 challenges in the inaugural GHIST competition in population genetics inference. He not only submitted accurate inferences, he also gamed the system by further optimizing his score beyond the maximum likelihood inferences. 😜