Open Jigyasa3 opened 9 months ago
We do not support running foldseek with amino acids at the moment. For that, foldseek would need to predict the structures, which we believe is currently outside the scope of the software. However, @milot-mirdita has developed a feature for our webserver based on the language model ProstT5. I believe ProstT5 allows fast inference of 3Di from an amino acid sequence. So you could build the workflow based on ProstT5 and foldseek.
Hi @martin-steinegger & @milot-mirdita
I have a single file with multiple 3Di sequences predicted from ProstT5, can you suggest how I can build a workflow to search against a custom build database? Also, I would like to know how to filter the final result in the output table similar to foldseek server's output.
Hi @martin-steinegger ,
Thank you for the awesome software for structure-based clustering and sequence search. I was wondering if its possible to run the
easy-search
oreasy-cluster
commands with protein sequences as input?