BioinformaticsbioRxivPreprint
✦ In-depth digest
Protein language models are biased by unequal sequence sampling across the tree of life
Frances Ding, Jacob Steinhardt
Likelihoods from Progen2 and ESM2 protein language models are systematically higher for sequences from some species, tracking evolution-weighted database sequence counts, and pull likelihood-guided designs away from extremophile properties.
57 citations