isGPT: An optimized model to identify sub-Golgi protein types using SVM and Random Forest based feature selection.

Area of Science:

• Molecular Biology
• Bioinformatics

Background:

• The Golgi Apparatus (GA) is crucial for modifying and sorting proteins in eukaryotic cells.
• Dysfunctional GA proteins are linked to congenital glycosylation disorders and diseases like diabetes, cancer, and cystic fibrosis.
• Accurate classification of cis-Golgi and trans-Golgi proteins is vital for drug development.

Purpose of the Study:

• To develop a novel computational model for classifying cis-Golgi and trans-Golgi proteins.
• To introduce efficient feature extraction methods from protein sequences.
• To optimize the classification of sub-Golgi proteins.

Main Methods:

• Feature extraction from protein sequences.
• Feature ranking using Random Forest (RF).
• Classification using Support Vector Machine (SVM) on selected features.
• Application of Synthetic Minority Over-sampling Technique (SMOTE) for imbalanced datasets.

Main Results:

• The developed model, identification of sub-Golgi Protein Types (isGPT), achieved high accuracy: 95.4% (10-fold cross-validation), 95.9% (jackknife test), and 95.3% (independent test).
• The regression model outperformed the classification model.
• isGPT demonstrated superior performance compared to existing state-of-the-art techniques.

Conclusions:

• The isGPT model offers an effective approach for classifying cis-Golgi and trans-Golgi proteins.
• Accurate protein classification can significantly contribute to advancements in drug development for associated diseases.
• The study provides open-source code and datasets for reproducibility and further research.