Clarification on "BindingLevel" in train_test_data_folds.xlsx

[underdog199]

Hi @banerjeeamitava,

First of all, thank you for this great work. I have a quick question regarding the BindingLevel column (NB, WB, SB) in the train_test_data_folds.xlsx file. (Link: https://github.com/meyer-lab-cshl/BATMAN-paper/blob/main/results_batman/tcr_epitope_datasets/mutational_scan_datasets/train_test_data_folds.xlsx)

Could you clarify whether these labels are derived from computational predictions or from experimental validation results?

Thank you very much, and I look forward to your response.

[underdog199]

I guess those are predictions from NetMHCpan 4.1. Please correct me if i'm wrong.