DAPredict

DAPredict predicts drug action phenotypes, including therapeutic actions and adverse drug reactions (ADRs), by integrating chemical-genomics and pharmacogenomics information to support drug discovery and repositioning.

Key Features:

• Extensive Predictive Relationships: Houses 305,981 predicted relationships linking 1,748 approved drugs with 454 ADRs and 1,478 approved drugs with 178 Anatomical Therapeutic Chemical (ATC) categories.
• Novel Prediction Algorithm: Implements an algorithm that merges chemical genomics and pharmacogenomics to predict phenotype-based drug actions from local active structures of drugs and proteins.
• Large-Scale Predictions: Enables prediction of action phenotypic spectra for over 110 million compounds, including approximately 168,000 natural products.
• Mechanism Exploration and Structural Optimization: Maps phenotype-corresponding domains and substructures to explore mechanisms of drug actions and to support structural optimization aimed at reducing serious adverse reactions and improving druggability.

Scientific Applications:

• Druggability Assessment: Evaluates the druggability of new compounds by predicting action phenotypes and potential ADRs.
• Drug Repositioning: Identifies potential new indications by comparing predicted phenotype spectra across approved drugs and compounds.
• Safety and Efficacy Evaluation: Supports analysis of potential mechanisms underlying ADRs and predicted therapeutic effects to inform safety and efficacy assessments.
• Structural Optimization: Guides structural modification by linking substructures and domains to phenotypes to mitigate adverse effects and enhance druggability.

Methodology:

Predicts drug action phenotypes using an algorithm that merges chemical genomics and pharmacogenomics and infers phenotypes from local active structures of drugs and proteins.