Method and systems for prediction of hla class ii-specific epitopes and characterization of cd4+ t cells

A method comprising:

• (a) processing amino acid information of a plurality of candidate peptide sequences using a machine learning HLA peptide presentation prediction model to generate a plurality of presentation predictions, wherein each candidate peptide sequence of the plurality of candidate peptide sequences is encoded by a genome or exome of a subject, or a pathogen or a virus in the subject, wherein the plurality of presentation predictions comprises an HLA presentation prediction for each of the plurality of candidate peptide sequences, wherein each HLA presentation prediction is indicative of a likelihood that one or more proteins encoded by a class II HLA allele of a cell of the subject can present a given candidate peptide sequence of the plurality of candidate peptide sequences,
• wherein the machine learning HLA peptide presentation prediction model is trained using training data comprising sequence information of sequences of training peptides identified by a method comprising:
• (A) incubating an HLA class II tetramer or multimer in the presence of (i) a soluble HLA-DM, (ii) a fluorescently labeled peptide probe, and (iii) a training peptide epitope loaded with an epitope with a HLA class II tetramer or multimer, thereby forming a first complex comprising the HLA class II tetramer or multimer and the fluorescently labeled peptide probe and a second complex comprising the HLA class II tetramer or multimer and the training peptide epitope;
• (B) measuring polarization;
• and (C) determining an affinity of the training peptide epitope to the HLA class II tetramer or multimer based on (B);
• and
• (b) identifying, based at least on the plurality of presentation predictions, a peptide sequence of the plurality of peptide sequences as being presented by at least one of the one or more proteins encoded by a class II HLA allele of a cell of the subject.