Lightnews — Scholar-powered news

Mike-ghoul Half-man

@michaelhoffman.bsky.social

16K followers 1.6K following 4.5K posts

Chair, Computational BIology and Medicine Program, Princess Margaret Cancer Centre, University Health Network. Associate Professor, Medical Biophysics, University of Toronto. Disclosures: https://github.com/michaelmhoffman/disclosure/

github.com

Posts Media Videos Starter Packs

Pinned

Mike-ghoul Half-man @michaelhoffman.bsky.social · Nov 12

I'm at Princess Margaret Cancer Centre and University of Toronto. We have three major interests:

1/Computational methods and #ML for gene regulation data 🧬🖥️
2/Liquid biopsy beyond simple genetic variants (epigenomic, immunogenomic)
3/Reproducibility and robustness of biomedical analysis (esp. #ML)

2 6 57

Mike-ghoul Half-man @michaelhoffman.bsky.social · 7m

FL: Placing the genetic architecture of human phenotypes into context. To predict phenotypes from genotypes we cannot ignore environmental or non-genetic effects. #ASHG25

Mike-ghoul Half-man @michaelhoffman.bsky.social · 8m

Now: Francesca Luca: Genetic risk and gene regulation across changing environments #ASHG25 🧪🧬🖥️

Mike-ghoul Half-man @michaelhoffman.bsky.social · 9m

LY: Larger studies of admixed individuals are needed. #ASHG25

Mike-ghoul Half-man @michaelhoffman.bsky.social · 10m

LY: Haplotype frequency differences are the main contributors to the loss of prediction accuracy between populatoins. #ASHG25

Mike-ghoul Half-man @michaelhoffman.bsky.social · 10m

LY: There are some but not many population-specific associations. Need more data from continental Africa. #ASHG25

Mike-ghoul Half-man @michaelhoffman.bsky.social · 11m

LY: Strong evidence of shared causal effects of common variants across traits. Admixture-based estimates are noisy for specific traits. #ASHG25

Mike-ghoul Half-man @michaelhoffman.bsky.social · 11m

LY: In 5–7 years we will build better statistical models that can predict across all populations. #ASHG25

Mike-ghoul Half-man @michaelhoffman.bsky.social · 12m

LY: Linkage disequilibrium and minor allele frequenceis differences between populations are the main contributors to the lack of transferability of polygenic scores. #ASHG25

Mike-ghoul Half-man @michaelhoffman.bsky.social · 21m

LY: Ancestry diversity in GWAS has improved greatly, especially for African-American and Afro-Carribean populations. #ASHG25

Mike-ghoul Half-man @michaelhoffman.bsky.social · 23m

Now: Loic Yengo: The genetic architecture of complex traits through the lens of multi-ancestry genetic studies #ASHG25 🧪🧬🖥️

1 3

Mike-ghoul Half-man @michaelhoffman.bsky.social · 24m

I didn't state this here. They did "realistic simulations" of various scenarios and their data did not match the stabilizing simulation scenario.

1 1

Mike-ghoul Half-man @michaelhoffman.bsky.social · 26m

DR: Our work focused on Europe in the last 10000 years. To what extent will patterns in other places and times be different? #ASHG25

Mike-ghoul Half-man @michaelhoffman.bsky.social · 26m

DR looks at predicted years of schooling polygenic score from 8000 years ago but obviously there was no school. Highly correlated with other traits like walking pace and age of first child. #ASHG25

Mike-ghoul Half-man @michaelhoffman.bsky.social · 28m

DR: Visualize selection information at 9.7 million variants and compare with UK Biobank traits #ASHG25

AGES

reich-ages.rc.hms.harvard.edu

Mike-ghoul Half-man @michaelhoffman.bsky.social · 29m

DR: Celiac disease risk rose (HLA-DQB1 rs3891176 C>A and TSBP1) in the last 4000 years. Selection reversal at TYK2 risk factor for severe tuberculosis. #ASHG25

Mike-ghoul Half-man @michaelhoffman.bsky.social · 31m

DR: Second answer: selection has fluctuated over time despite assumptions usually used to simplify analyses. #ASHG25

Mike-ghoul Half-man @michaelhoffman.bsky.social · 31m

DR: Evidence of acceleration in selection over last 5,000 years. #ASHG25

Mike-ghoul Half-man @michaelhoffman.bsky.social · 31m

DR: How to reconcile our result with lack of evidence for selective sweeps? Answer: we are in a period of intensified evolution in the last 10,000 years. Immune traits enriched for selection signals 5.7×, metabolic 2.1×, but none for psychiatric or behavior. #ASHG25

1 1

Mike-ghoul Half-man @michaelhoffman.bsky.social · 33m

DR: Alleles with signals of selection are enriched for alleles affecting traits in 452 genome-wide association studies. This is only expected for directional selection, not stabilizing, balancing, or purifying selection. #ASHG25

Mike-ghoul Half-man @michaelhoffman.bsky.social · 34m

DR: A common view is that selection was quiescent in the last 10000 years. There are a small number of loci with evidence of selection #ASHG25

Mike-ghoul Half-man @michaelhoffman.bsky.social · 38m

Now: David Reich: Pervasive findings of directional selection realize the promise of ancient DNA to elucidate human adaptation #ASHG25 🧪🧬🖥️

1 4

Mike-ghoul Half-man @michaelhoffman.bsky.social · 40m

JK: Allele of the EPAS1 gene gives Tibetans better tolerance of high altitudes. Genetic analysis showed it was contributed by Denisovans. Then archaeologists found Denisovan material in a cave on the Tibetan plateau #ASHG25

Mike-ghoul Half-man @michaelhoffman.bsky.social · 40m

Mike-ghoul Half-man @michaelhoffman.bsky.social · 45m

JK: Three Toll-like receptors on chromosome 4: TLR10, TLR1, TLR6. Neandertal introgressed alleles have higher expression of these genes, don't affect the protein #ASHG25

2 1

Mike-ghoul Half-man @michaelhoffman.bsky.social · 46m

JK: Adaptive introgression in modern humans. Consistently linked to systems involved in interaction with the environment: immune genes, metabolic processes, environmental adaptations (latitude, light exposure). Came from people who lived in these areas for hundreds of thousands of years #ASHG25