Title: Computational Gene Expression Analysis in HIV and HIV/TB Co-Infected Individuals of Microarray Data by Employing Network Pharmacology and Machine Learning Drug Designing Methodologies
Keywords: TB/HIV, Coinfection, Microarray analysis, Network biology, QSAR-ML, Virtual screening, molecular docking.
Description:
Identification of a therapeutic target through microarray datasets for the TB-HIV coinfection state. The identified target was further explored using cheminformatics and machine learning to identify potent inhibitors capable of modulating the coinfection state.
Methodology:
-
Bioinformatics and Network Analysis:
Gene expression analysis was performed on TB-HIV coinfected patient samples. Highly expressed genes were utilized to construct a protein-protein interaction (PPI) network, which was analyzed using various network descriptors. The analysis prioritized the most central and high-degree gene, identified as EIF4A3. -
Cheminformatics and Machine Learning Modeling:
A robust QSAR model for EIF4A3 classification was developed. The best-performing model was employed for virtual screening of anti-TB and HIV chemical libraries. The top-ranked molecules were further evaluated through structure-based approaches, drug-likeness assessment, and metabolite analysis, ultimately leading to the identification of two probable hits.