AI summaryⓘ
The authors studied how well different models predict how strongly peptides bind to proteins by combining several datasets. They tested various ways to represent peptides and proteins, as well as different machine learning methods, across three types of data splits to see if models truly generalize or just memorize. Their results showed that choice of model and representation mattered most when training data was small, but differences decreased with more data. They also found that adapting some advanced models did not consistently improve performance. Overall, they suggest that evaluating peptide-protein binding models needs careful data splitting based on the intended use and consideration of data size and model types.
Peptide-protein bindingMachine learning regressionData partitioningMolecular representationSpearman correlationKd Ki IC50Random forestESM-2 embeddingsModel generalization
Abstract
Peptide-protein affinity models are often evaluated with a single data split, obscuring whether they interpolate among measurements for observed targets or generalize across peptide or target shifts. We integrated three sources of quantitative peptide-protein binding data to obtain 11,349 deduplicated pairs and benchmarked ten peptide representations, ESM-2 protein embeddings, and six regressors under peptide-similarity, within-target, and leave-target-out partitions. Across 60 matched representation-regressor configurations, mean test Spearman correlations were 0.462, 0.669, and 0.530, respectively. The top configuration shifted from ECFP-16 count fingerprints with random forest in the first two settings to HELM-BERT with Extra Trees when exact target sequences were excluded. Representation-rank correlations ranged from -0.042 to 0.624 across partitions, whereas regressor-rank correlations ranged from 0.771 to 0.943. Learning curves showed that representation differences were largest with limited supervision and narrowed as training data increased. PeptideCLM-2 adaptation and simple element-wise interaction features provided no consistent gain over a frozen encoder and direct concatenation under the tested protocols. These conclusions are specific to a dataset that pools transformed Kd, Ki, and IC50 measurements and to target exclusion at the exact-sequence level. Peptide-protein affinity benchmarks should therefore align data partitions with the intended use and jointly assess the effects of data scale, molecular representation, and downstream learner.