Seyone Chithrananda

Post-preprint, we’ve added guidance on how to check for exchangeability (i.e.: does a given similarity threshold’s guarantee hold on your test proteins of interest?) prior to calibration, refined code usability and notation.

It was awesome working on this with Ron, Anastasios, & crew! (3/3)

Our work provides steps for how to do calibrated annotation for domains of unknown function, enzyme search, and to conduct proteome-scale structural searches in a fraction of the time through pre-filtering.

Think of it as a “model-free pLDDT”-style confidence for embedding-based search! (2/3)

Excited to share our work on calibrated protein search and genome mining with conformal prediction, now in @naturecomms.bsky.social!
📎 rdcu.be/d5pJG
💻 github.com/ronboger/confo…
(1/3)

Ultrafast classical phylogenetic method beats large protein... Amino acid substitution rate matrices are fundamental to statistical phylogenetics and evolutionary biology. Estimating them typically requires reconstructed trees for massive amounts of aligned...

Large protein language models can learn complex epistatic interactions, but how much does that help with predicting variant effects? In this NeurIPS article, we show that classical independent-sites phylogenetic models can outperform pLMs on this task.
1/7
openreview.net/forum?id=H7m...