GCAC

GCAC performs predictive model building for compound activity classification and virtual screening using descriptor calculation, feature selection, and statistical modeling within R and Galaxy workflows.

Key Features:

• Integration with Galaxy workflow system: Implements reproducible workflows via the Galaxy workflow system to ensure transparent, repeatable analysis steps.
• Implementation in R: Uses the R Statistical Computing Environment and its statistical packages for descriptor calculation, feature selection, and model fitting.
• Data mining pipeline for high-throughput virtual screening (HTS): Provides a pipeline for mining and processing large HTS and virtual screening compound datasets.
• Descriptor calculation and feature selection: Calculates molecular descriptors and selects relevant features for model input using R statistical packages.
• Predictive model building and virtual screening: Builds statistical predictive models to classify compound activity and screens large compound libraries to prioritize candidates.
• Automated workflow configuration: Automates configuration settings within Galaxy workflows to standardize analysis parameters.

Scientific Applications:

• Cheminformatics: Enables descriptor-based analysis and feature selection for molecular property and activity studies.
• Drug discovery and virtual screening: Prioritizes compounds from large libraries using predictive models to support hit identification and lead selection in drug discovery pipelines.

Methodology:

Computational steps explicitly include molecular descriptor calculation, feature selection, building statistical predictive models in R to predict compound activities, and applying those models for virtual screening of large compound datasets.