Neoantigen × AI
Daily research signal
№59Brief

DeepOmicsNeo improves MHC class I epitope prediction

Today's brief is a single method signal: a bioRxiv preprint, DeepOmicsNeo, reports improved MHC class I epitope prediction by attacking the field's core data problems — scarce binding data for non-9mer peptides and the absence of true negatives for immunogenicity.bioRxiv

Epitope selection is the highest-leverage step in personalized neoantigen vaccines, so predictor gains translate directly into target-list quality and per-patient economics. As with all preprints, the benchmark claims await independent and prospective validation.

1new story
AI / MethodsyesterdayNew

Drastic improvement in MHC class I epitope prediction through training data augmentation, generative adversarial network modeling of selfness, and consideration of homology to class II epitopes.

A bioRxiv preprint introduces DeepOmicsNeo, a modular deep learning framework combining three components: a binding module that synthesizes pseudo-binders and non-binders by editing non-anchor residues to overcome data scarcity in non-9mer peptide–MHC class I prediction; a 'selfness' module using a GAN trained on approximately 60 million human self-peptides to quantify similarity to self in an unsupervised manner, with scores correlating with IFN-gamma secretion by T cells; and a module exploiting central-region homology between HLA class I and class II epitopes. The integrated immunogenicity model reportedly outperforms state-of-the-art predictors in benchmark testing. Why it matters: high false-positive rates in current predictors are a key rate limiter for personalized cancer vaccines, and a better-ranking immunogenicity model could raise the hit rate of selected targets without adding clinical cost. It is method-stage and unreviewed — no prospective or clinical validation is reported.

biorxiv · yesterday · Oh, S.; Hong, S.-E.; Kim, H.-S.; Paik, S.