Output File and Rerunning the Model

[muradrmamedov]

Dear Dr. Cong and colleagues,

I was very excited to read your recent bioRxiv manuscript, “Computing the Human Interactome.” I was hoping you could address a couple of my questions. Firstly, what exactly is reported in the output tables report that is generated by RoseTTAFold2-PPI (outputs in https://github.com/CongLabCode/RoseTTAFold2-PPI/tree/main/examples/expected_output)? Secondly, I am interested in PPI identification for specific protein families. If we reran your trained model on a subset of proteins, would we find substantially different predictions than the predictions already provided in your manuscript and website (http://prodata.swmed.edu/humanPPI/browse)?

Thank you,
Murad Mamedov