THETA: a new genotypic approach for predicting HIV-1 CRF02-AG coreceptor usage

Area of Science:

• Virology
• Bioinformatics
• Computational Biology

Background:

• The circulating recombinant form of HIV-1 CRF02-AG is a prevalent non-B subtype in Europe.
• Accurate determination of HIV-1 tropism is crucial for anti-HIV therapy and understanding disease progression.
• Existing genotypic tropism prediction methods, V3-based, show poor performance for CRF02-AG.

Purpose of the Study:

• To develop a more accurate genotypic tropism prediction algorithm for HIV-1 CRF02-AG.
• To improve the reliability of tropism determination for non-B HIV-1 subtypes.

Main Methods:

• Designed a novel algorithm, Toulouse HIV Extended Tropism Algorithm (THETA), utilizing bio-statistical methods.
• Employed a Least Absolute Shrinkage and Selection Operator (LASSO) approach on HIV envelope sequences (gp120 and gp41 ectodomain) from phenotypically characterized clones.
• Validated the algorithm's predictive performance.

Main Results:

• THETA achieved 86% sensitivity for predicting R5X4/X4 viruses and 89% specificity for R5 viruses.
• The overall accuracy of the THETA algorithm was 88% for predicting HIV-1 CRF02-AG tropism.
• The developed algorithm demonstrates sufficient reliability for subtype CRF02-AG.

Conclusions:

• THETA provides a reliable method for genotypic tropism prediction of HIV-1 CRF02-AG.
• This algorithm enhances the ability to guide anti-HIV therapy and conduct pathophysiological studies for this specific HIV-1 subtype.
• The tool is freely available, facilitating wider application in clinical and research settings.