Accurate Identification of Human Phosphorylated Proteins by Ensembling Supervised Kernel Self-organizing Maps

Area of Science:

• Biochemistry
• Bioinformatics
• Computational Biology

Background:

• Protein phosphorylation is a critical post-translational modification regulating numerous cellular processes.
• Accurate prediction of protein phosphorylation from sequence is essential for biological research and therapeutic development.

Purpose of the Study:

• To develop a novel, sequence-based predictor for identifying human phosphorylated proteins.
• To improve the accuracy and efficiency of predicting protein phosphorylation compared to existing methods.

Main Methods:

• Developed a predictor using two supervised kernel self-organizing maps (SKSOMs).
• One SKSOM utilized protein physiochemical composition features, while the other used evolutionary information.
• Ensembled the predictions from both SKSOMs for the final classification.

Main Results:

• Achieved an accuracy (ACC) of 78.75% and a Matthews Correlation Coefficient (MCC) of 0.561.
• Demonstrated significant improvements over the state-of-the-art predictor, with ACC and MCC increases of 6.96% and 12.5%, respectively.

Conclusions:

• The proposed method offers a sensitive and accurate approach for identifying human phosphorylated proteins.
• This computational tool has potential applications in basic research and drug discovery.
• The methodology can be extended to predict phosphorylation in proteins from other species.