Claude can predict the next amino acid of proteins nobody has ever studied

Claude was trained on protein sequences at scale

Claude can predict the next amino acid of proteins nobody has ever studied

Give the latest Claude models the start of an obscure bacterial protein and they guess the next amino acid right 54% of the time; shuffle the sequence and accuracy collapses to 9%. GPT, Gemini, and seven open-weight models stay near zero on the same 668 proteins from 167 Pfam families of unknown function. Accuracy climbs with family size even in families with no papers, pointing to large-scale protein sequence data in training rather than reasoning or literature.

The latest Claude models know protein families that almost nobody has studied. That takes a lot of sequence data, not a few examples: protein sequences were a sizeable part of their training.
  1. eddiewithzato

    Claude just blocks from anything in this field. Can't even work with my own DNA without triggering opus 5.5 50% of the time

  2. consumer451

    https://wallhop.io/http://xinmingtu.cn/blog/2026/claude-opus...

  3. zodwick

    Is this a valuable skill ? Sorry, bit of a noob question ig but have been seeing a lot of buzz around protein sequences in general

  4. pcrh

    It's hard to see what the point of this is?

    When would one want to predict a full-length unknown protein sequence from only a fragment? Any fragment of any useful length can be back-translated to the DNA sequence and that can be tested against genomic databases (a process called tblastn). And if your protein/gene is not found there, you have no way of knowing whether or not the predicted sequence has any basis in reality.

    Testing the model against shuffled sequences is also kinda cheating, as these are no longer random pool of amino acids, but one whose source already exists in the databases.

  5. octoberfranklin

    Curious: how can there be a correct prediction for the next element of a "shuffled" protein? The shuffled proteins don't occur in real life, so what makes one answer correct and another incorrect?

More from this day

2026-11-18