PHOSFER

PHOSFER predicts phosphorylation sites in plant proteins, with an initial focus on soybean, to identify regulatory post-translational modification sites.

Key Features:

• Random forest model: Uses a random forest-based machine-learning classifier for phosphorylation site prediction.
• Cross-species data integration: Leverages and integrates phosphorylation data from multiple organisms to improve generalization across plant species.
• Soybean focus: Initially trained and evaluated with emphasis on soybean proteins.
• Comparative performance: Demonstrated superior performance relative to Arabidopsis-specific predictors and simpler models using only known soybean phosphorylation sites.

Scientific Applications:

• Soybean phosphoproteomics: Support identification of phosphorylation sites in soybean proteins for soybean research.
• Plant signaling and regulation studies: Facilitate analysis of phosphorylation dynamics and regulatory post-translational modifications in plant biology.
• Cross-species prediction: Enable extension of phosphorylation site predictions to other plant species through integrated cross-species data.

Methodology:

PHOSFER trains a random forest machine-learning model on integrated phosphorylation datasets compiled from multiple organisms.