UPFPSR

UPFPSR predicts lysine ubiquitylation sites in plant proteins to identify post-translational modification locations relevant to regulation of cellular processes.

Key Features:

• Feature representation: Uses multiple physicochemical properties of amino acids and sequence-based statistical information as input features for prediction.
• Classification algorithm: Employs a random forest classifier selected from an evaluation of four traditional algorithms and two deep learning networks.
• Predictive performance: Achieves 77.3% accuracy, 75% precision, 81.7% recall, 0.7824 F1-score, and 0.84 AUC on independent test datasets.
• Benchmarking: Demonstrates superior performance across multiple measurement indicators in comparisons with existing ubiquitylation prediction tools.

Scientific Applications:

• Site identification: Enables rapid in silico identification of lysine ubiquitylation sites in plant proteomes.
• Experimental prioritization: Guides selection of candidate sites for experimental validation of ubiquitylation.
• Functional studies: Supports investigation of ubiquitylation-mediated regulation of cellular processes and pathways in plants.

Methodology:

Combines multiple physicochemical amino acid properties and sequence-based statistical features and applies a random forest classifier chosen from four traditional algorithms and two deep learning networks.