Contrastive learning unites sequence and structure in a global representation of protein space
Proceedings of the National Academy of Sciences, Volume 123, Issue 32, August 2026. SignificanceHow can the relationships between all known proteins be represented? Recent advances in AI describe proteins as vectors, where the distance between vectors relates to the similarity between proteins. These vector representations are largely ...
We haven't written up this one. PNAS has the full story — the link below goes straight to it.
