Zero-shot performance of PLMs, but not inverse folding models, correlates with number of homologs available for training

Summary

Zero-shot fitness prediction performance with protein language models, but not structure-based models (e.g., inverse folding models), correlates with number of homologs available for training (1).

Figures

Ref (1)

See also

• ESM-IF, but not other inverse folding models, has learned some evolutionary constraints from sequence databases

1.

Li F-Z, Yang J, Johnston KE, Gürsoy E, Yue Y, Arnold FH. Evaluation of machine learning-assisted directed evolution across diverse combinatorial landscapes. Cell Systems. 2025;16(9):101387. Available from: https://doi.org/10.1016/j.cels.2025.101387