GIpred: a computational tool for prediction of GIGANTEA proteins using machine learning algorithm.

Area of Science:

• Plant molecular biology
• Computational biology
• Bioinformatics

Background:

• GIGANTEA (GI) proteins regulate vital plant processes like metabolism and growth.
• Experimental identification of GI proteins is time-consuming and resource-intensive.

Purpose of the Study:

• To develop a fast and accurate computational model for predicting GI proteins.
• To facilitate large-scale identification and functional annotation of GI proteins.

Main Methods:

• Employed ten supervised learning algorithms (SVM, RF, etc.).
• Utilized amino acid composition (AAC), FASGAI features, and physico-chemical (PHYC) properties as input.
• Validated models using five-fold and leave-one-out cross-validation.

Main Results:

• Achieved high prediction accuracies: up to 97.29% AUC-ROC and 87.89% AUC-PR.
• The SVM model with AAC+PHYC features showed superior performance.
• Successfully predicted 17 out of 18 GI sequences in an independent test dataset.

Conclusions:

• The developed computational model, "GIpred," offers an efficient method for GI protein prediction.
• Facilitates proteome-wide identification and functional annotation of GI proteins, particularly in crops like wheat.
• Provides a freely accessible online server for broader scientific use.