On structural benchmarks, we found that our model's structural representations outperform those from existing sequence models and even match performance with representations from models that take structure and sequence as input. (5/7)

ISM's secret sauce is a microenvironment-based autoencoder. The all-atom autoencoder learns to embed the tertiary structure surrounding a residue into a structure token. We distill these per-residue tokens (and MutRank tokens) into ESM2. (4/7)

Are you using ESM2 for your sequence embeddings? Try out ISM, a one-line code change that will incorporate improved structure and sequence information, without a structure as input. (1/7)