end-to-end PPI affinity prediction

[ChenShengsGitHub]

Great work! But I wonder whether your method support end-to-end PPI affinity prediction (i.e., input a pair of protein sequence and output PPI affinity)?

[ChenShengsGitHub]

I mean, for example, in your peptide meatadata, for per peptide, there are four element like "P26651,SESGRCRyGAKCQFA,phosphosite,1”. If I don't know the peptide ID and the "phosphosite,1", I only know the peptide sequence and protein sequence, can your method work in my task to end-to-end predict the PPI affinity?