search

Graylab / IgFold

Fast, accurate antibody structure prediction from deep learning on massive set of natural antibodies
Other
327 stars 61 forks source link

A question about benchmark datasets #11

Closed YT-bomb closed 2 years ago

YT-bomb commented 2 years ago

Recently, I have tried to reproduce the evaluation of Igfold, but have not found the download location of benchmark datasets. I tried to use the method mentioned in the paper, but the number of datasets is much larger than that in the paper. So I want to know whether you can provide relevant data ?Thanks a lot.

jeffreyruffolo commented 2 years ago

You can find the benchmark results here. Please note that there have been a few updates to the repository since they initial benchmarking, so results may vary slightly. I am working on an updated benchmark with the new structures that have been released since we compiled this set.

jiaxiang-wu commented 2 years ago

Hi @jeffreyruffolo

Thanks for providing these benchmark results. I noticed that only IgFold-predicted PDB files, together with evaluation results are included. Could you please also add native PDB files used in your experiments, so that we can run evaluation with PDB files predicted by other pipelines? I've tried with the original PDB files downloaded from RCSB, but there were some PyRosetta-related errors when directly using them for evaluation.

Best,