UPFPSR: a ubiquitylation predictor for plant through combining sequence information and random forest

Area of Science:

• Biochemistry
• Computational Biology
• Plant Science

Background:

• Ubiquitylation is a crucial post-translational modification regulating vital eukaryotic cellular processes.
• Traditional methods for identifying ubiquitylation sites are laborious and time-consuming.

Purpose of the Study:

• To develop an automated computational tool for rapid and accurate prediction of lysine ubiquitylation sites in plants.
• To enhance the efficiency of experimental studies on ubiquitylation.

Main Methods:

• Developed UPFPSR using physicochemical properties and sequence-based statistical information.
• Compared four traditional algorithms and two deep learning networks, selecting random forest for optimal performance.
• Validated performance on an independent test dataset.

Main Results:

• UPFPSR achieved high performance metrics: 77.3% accuracy, 75% precision, 81.7% recall, 0.7824 F1-score, and 0.84 AUC.
• Outperformed existing advanced ubiquitylation prediction tools on all measurement indicators.

Conclusions:

• UPFPSR provides a novel, efficient, and accurate computational approach for predicting plant ubiquitylation sites.
• The tool can guide future experimental investigations into ubiquitylation processes in plants.