1 Prediction of Protease Substrates Using Sequence and Structure Features

Motivation: Granzyme B (GrB) and caspases cleave specific pro-tein substrates to induce apoptosis in virally infected and neoplastic cells. While substrates for both types of proteases have been de-termined experimentally, there are many more yet to be discovered in humans and other metazoans. Here, we present a bioinformatics method based on support vector machine (SVM) learning that identi-fies sequence and structural features important for protease recogni-tion of substrate peptides and then uses these features to predict novel substrates. Our approach can act as a convenient hypothesis generator, guiding future experiments by high confidence identifica-tion of peptide protein partners. Results: The method is benchmarked on the known substrates of both protease types, including our literature-curated GrB substrate set (GrBah). On these benchmark sets, the method outperforms a number of other methods that consider sequence only, predicting at a 0.87 true positive rate and a 0.13 false positive rate for caspase substrates, and a 0.79 true positive rate and a 0.21 false positive rate for GrB substrates. The method is then applied to ~25,000 pro-teins in the human proteome to generate a ranked list of predicted substrates of each protease type. Two of these predictions, AIF-1 and SMN1, were selected for further experimental analysis, and each was validated as a GrB substrate. Availability: All predictions for both protease types are publically available a