Related Experiment Videos
Structure-based phenotyping predicts HIV-1 protease inhibitor resistance.
Mark D Shenderovich1, Ron M Kagan, Peter N R Heseltine
1Cengent Therapeutics Inc., 10929 Technology Place, San Diego, CA 92127, USA. marksh@cengent.com
Protein Science : a Publication of the Protein Society
|July 24, 2003
Summary
A new computational method accurately predicts HIV-1 drug resistance by analyzing binding energy changes in protease variants. This structural phenotyping approach offers a rapid and reliable alternative to traditional resistance assays.
Area of Science:
- Computational Biology
- Structural Biology
- Virology
Background:
- HIV-1 drug resistance arises from mutations in drug targets, leading to treatment failure.
- Current phenotypic resistance assays are slow and expensive, while genotypic interpretations can be unreliable with multiple mutations.
Purpose of the Study:
- To develop a rapid computational method for evaluating HIV-1 drug resistance.
- To assess the accuracy of a structure-based approach in predicting resistance to protease inhibitors (PIs).
Main Methods:
- Computational modeling of wild-type (WT) and mutant HIV-1 protease (PR) complexes with PIs.
- Calculation of binding energy changes (DeltaE(bind)) between mutant and WT complexes.
- Correlation of calculated DeltaE(bind) with experimentally determined phenotypic resistance (IC(50) ratios).
Main Results:
- Calculated DeltaE(bind) showed significant correlations with IC(50) ratios from phenotypic assays (R(2) = 0.7-0.85).
- A structure-based phenotype prediction achieved 92% agreement with cell-based assays for 78 PR variants.
- The method accurately predicted drug resistance in clinical HIV-1 PR variants.
Conclusions:
- Computational structural phenotyping is a rapid and accurate method for predicting HIV-1 drug resistance.
- This approach can complement or potentially replace traditional phenotypic assays.
- The developed method shows high accuracy in predicting resistance profiles of clinical HIV-1 PR variants.