Clustering ProstT5 DB?

[HaimAshk]

Dear FoldSeek team,

Sorry for maybe a very novice question, I could not find the answer in the documentation...
I have a ProstT5-encoded DB for foldseek. Is there an easy way to cluster all the proteins included in the DB? It seems easy-cluster scheme does not accept a DB as an input (or I missed it).
How can I achieve it?

Thanks for any advice!
Haim

[milot-mirdita]

That should work. What did you execute? You might need to use cluster instead of easy-cluster since the latter only accepts structures and not databases as input

[HaimAshk]

Thanks for the prompt response, Milot!
Yes, using just cluster seems to work. Thanks!

[shiraz-shah]

@HaimAshk, could you please share the commands as to how you succeded with this? I have prostT5 3di fastas for my proteins and can't get cluster to work for the life of me. It just throws a database type error.

[milot-mirdita]

easy-cluster will call ProstT5 internally if you give it a FASTA file. cluster cannot do this.