ApoplastP: prediction of effectors and plant proteins in the apoplast using machine learning

Abstract The plant apoplast is integral to intercellular signalling, transport and plant-pathogen interactions. Plant pathogens deliver effectors both into the apoplast and inside host cells, but no computational method currently exists to discriminate between these localizations. We present ApoplastP, the first method for predicting if an effector or plant protein localizes to the apoplast. ApoplastP uncovers features for apoplastic localization common to both effectors and plant proteins, namely an enrichment in small amino acids and cysteines as well as depletion in glutamic acid. ApoplastP predicts apoplastic localization in effectors with sensitivity of 75% and false positive rate of 5%, improving accuracy of cysteine-rich classifiers by over 13%. ApoplastP does not depend on the presence of a signal peptide and correctly predicts the localization of unconventionally secreted plant and effector proteins. The secretomes of fungal saprophytes, necrotrophic pathogens and extracellular pathogens are enriched for predicted apoplastic proteins. Rust pathogen secretomes have the lowest percentage of apoplastic proteins, but these are highly enriched for predicted effectors. ApoplastP pioneers apoplastic localization prediction using machine learning. It will facilitate functional studies and will be valuable for predicting if an effector localizes to the apoplast or if it enters plant cells. ApoplastP is available at http://apoplastp.csiro.au .