Classification and regression trees--studies of HIV reverse transcriptase inhibitors

Area of Science:

• Medicinal Chemistry
• Computational Biology
• Drug Discovery

Background:

• Non-Nucleoside Reverse Transcriptase Inhibitors (NNRTIs) are crucial in HIV treatment.
• Understanding structure-activity relationships is key to developing more effective NNRTIs.
• HIV drug resistance necessitates analysis of drug activity against mutant strains.

Purpose of the Study:

• To apply Classification And Regression Trees (CART) for analyzing NNRTI biological activity.
• To correlate NNRTI interaction energies with the Reverse Transcriptase (RT) binding pocket to their observed activity.
• To elucidate the molecular determinants of NNRTI efficacy against wild-type and mutant HIV-1.

Main Methods:

• Utilized Classification And Regression Trees (CART) for predictive modeling.
• Analyzed biological activity data (pIC50) for 208 NNRTIs against HIV-1 and four mutant strains.
• Incorporated computed interaction energies between NNRTIs and the RT binding pocket.

Main Results:

• CART successfully identified key amino acid interactions driving NNRTI biological activity.
• The model explained variations in pIC50 values based on interactions with specific RT binding pocket residues.
• Biological activity was effectively predicted using interaction energies and CART analysis.

Conclusions:

• CART is a valuable tool for understanding NNRTI structure-activity relationships.
• Specific amino acid interactions within the RT binding pocket significantly influence NNRTI efficacy.
• This approach aids in the rational design of novel NNRTIs against drug-resistant HIV strains.