DrivR-Base

DrivR-Base extracts and integrates molecular features related to single nucleotide variants to support the construction of machine learning models for variant effect prediction.

Key Features:

• Genomic and protein position features: Extracts information about genomic positions and associated protein positions of variants.
• Structural property integration: Incorporates structural properties obtained from AlphaFold.
• Regulatory information integration: Includes regulatory information sourced from ENCODE.
• Predicted variant consequences: Integrates predicted variant consequences from Variant Effect Predictor.
• Docker containerization: Deployable via a Docker container to ensure reproducibility across diverse computational environments.

Scientific Applications:

• Pathogenic impact prediction: Provides features as input for machine learning models designed to predict the pathogenic impact of human genome variants in disease.
• Haploinsufficiency prediction: Feature sets can be applied to haploinsufficiency prediction tasks.
• Drug repurposing tool development: Features support the development of drug repurposing tools.

Methodology:

DrivR-Base extracts and integrates molecular features from multiple databases and tools including AlphaFold, ENCODE, and Variant Effect Predictor.