MHC II binding predictions

Predictions of peptide (15-mer) bindings to mouse I-A^b MHCII molecule using the neural network algorithm implemented in the IEDB MHC II version 2.13 collection of tools [1]. This data accompanies [2].Naming convention for results from microbial 15-mer binding is "mhciipred.[microbe_name]%[accession_number].txt".uniprot-proteome.3AUP000000589.fasta is the mouse proteome used for this study (available elsewhere, but included here for convenience and to avoid confusion with other proteome versions), while uniprot-proteome_random.3AUP000000589.fasta is a synthetic mouse proteome generated by randomly replacing each residue using a probability distribution that matches the amino acid usage in the original proteome.mhciipred.mouse.UP000000589.txt and uniprot-proteome_random.3AUP000000589.fasta, respectively, contain the binding predictions for the original and random mouse proteomes.[1] Nielsen, M. &amp; Lund, O. NN-align. An artificial neural network-based alignment algorithm for MHC class II peptide binding prediction. <i>BMC bioinformatics</i> <b>10</b>, 296 (2009).[2] Nettersheim, F.S., Brunel, S., Sinkovits, R.S., Armstrong, S.S., Roy, P., Billitti, M., Kobiyama, K., Alimadadi, A., Bombin, S., Lu, L., Zoccheddu, M., Oliaeimotlagh, M., Benedict, C.A., Sette, A. and Ley, K. PD1 and CD73 on naïve CD4+ T cells synergistically limit responses to self. Nature Immunology (2024) in press.