AI-assisted HIV drug resistance interpretation integrating validated rule-based logic with calibrated machine-learning models. This platform developed by the Integrated Computational Viromics and Digital Analyitics in collaboration with Ethiopian Public Health Institute
Interprets HIV-1 sequences across all major antiretroviral classes—protease inhibitors (PIs), nucleoside reverse transcriptase inhibitors (NRTIs), non-nucleoside reverse transcriptase inhibitors (NNRTIs), and integrase strand transfer inhibitors (INSTIs)—by analyzing protease (PR), reverse transcriptase (RT), and integrase (IN) regions for complete resistance profiling.
Provides resistance classifications aligned with Stanford HIVDB categories, delivering consistent Susceptible, Low, Intermediate, and High resistance levels for clinical research, surveillance, and expert consultation.
Supports batch sequence analysis with exportable CSV and JSON outputs, enabling scalable surveillance, programmatic monitoring, and research applications.
PGX delivers HIV drug resistance interpretation through two complementary analytical platforms:
DR-Genotype Portal: Uses curated mutation scoring aligned with WHO, Stanford HIVDB, and ANRS guidelines. Analyzes HIV-1 sequences for known resistance mutations across PR, RT, and IN genes, applying clinically validated scoring algorithms to generate standardized resistance levels (Susceptible, Low, Intermediate, High) for each antiretroviral drug. This transparent, rules-driven approach provides clinically interpretable outputs that align with established treatment guidelines.
Drug-Specific Insights: Machine learning-based phenotypic prediction using ensemble models trained on >45,000 genotype-phenotype pairs. Examines resistance patterns, mutation effects, and score distributions for individual antiretroviral drugs. Models are validated against independent phenotypic datasets and calibrated against PhenoSense distributions, providing quantitative resistance predictions with confidence intervals.
Key Integration: Both platforms can be used independently or in conjunction, delivering outputs that combine the clinical interpretability of rule-based systems with the predictive power of machine learning for comprehensive resistance assessment.
Navigate to specialized analytical modules for deeper insight
Upload HIV sequences to receive comprehensive resistance interpretations using validated mutation scoring algorithms aligned with clinical guidelines.
Examine resistance patterns, mutation effects, and score distributions for individual antiretroviral drugs using machine learning models.
Explore full training, validation, calibration, and robustness analyses for all machine-learning models.