EffectorP: Using Machine Learning to Predict Fungal Effector Proteins and Their Subcellular Localization in the Plant Cell
J. SPERSCHNEIDER (1), D. Gardiner (1), P. Dodds (1), K. Singh (1), J. Manners (1), J. Taylor (1) (1) CSIRO, Australia

Eukaryotic filamentous plant pathogens secrete effector proteins that modulate the host cell to facilitate infection. Computational effector candidate identification and subsequent functional characterization delivers valuable insights into plant-pathogen interactions. We previously developed EffectorP which uses machine learning to predict fungal effectors from secretomes based on a robust signal of sequence-derived properties, achieving sensitivity and specificity of over 80%. Of particular interest in the candidate effector set are cytoplasmic effectors that can enter the plant cell and might target specific cell compartments such as chloroplasts or nuclei to exhibit their virulence functions. However, accurate computational prediction of subcellular localization of cytoplasmic effectors in the plant cell has been challenging. Here, we introduce a novel machine learning classifier that can predict if a candidate effector carries a chloroplast/mitochondrial transit peptide or nuclear localization signal that can target it to a specific plant cell compartment. We find that our classifier shows higher accuracy on rust effectors as well as on oomycete effectors with published localization data than existing methods. Secretome-wide screening of effector candidates reveals that rust fungi might have evolved a class of effector proteins that translocate into chloroplasts and nuclei by mimicking plant-targeting sequences.

Abstract Number: S4-3
Session Type: Special Session