ActiveDriverDB

ActiveDriverDB annotates human genetic variation by mapping variants to post-translational modification (PTM) sites and signaling networks to interpret functional impacts in inherited disease and cancer.

Key Features:

• Extensive PTM Data Integration: Integrates over 260,000 experimentally validated PTM sites, including phosphorylation, glycosylation, sumoylation, and succinylation, with a 2021 update increasing coverage by ~50%.
• Machine Learning Prioritization: Applies machine learning algorithms to prioritize proteins and pathways affected by PTM-specific amino acid substitutions and induced or disrupted kinase-binding short linear motifs.
• Variant-to-Network Mapping: Maps genetic variants to site-specific protein interaction networks and known drug targets to predict consequences for cellular signaling pathways.
• Disease and Cancer Impact Quantification: Estimates that approximately 16–21% of amino acid substitutions among pathogenic disease mutations, somatic cancer mutations, and germline variants impact PTM sites.
• SARS-CoV-2 Phosphoproteomics Integration: Incorporates a phosphoproteomics dataset reflecting cellular responses to SARS-CoV-2 to analyze how human genetic variation may modulate infection responses.
• Functional Impact Examples: Links specific mutations to phenotypes, including CFTR mutations in cystic fibrosis and CTNNB1 phosphodegron alterations in cancer.

Scientific Applications:

• Genetic Variant Interpretation: Prioritizes functionally significant variants among millions of mapped genetic variants by assessing PTM impact.
• Disease Mechanism Elucidation: Dissects molecular mechanisms of inherited diseases and oncogenesis through PTM site analysis.
• Drug Target Discovery: Identifies potential drug targets by mapping variant effects onto protein interaction networks and drug-target sites.
• COVID-19 Research: Predicts how human genetic variation may influence SARS-CoV-2 infection and disease progression using phosphoproteomics data.
• Personalized Medicine: Informs personalized therapeutic strategies by linking individual variant impacts on PTMs and signaling pathways to potential interventions.

Methodology:

Integrates experimentally validated PTM sites and phosphoproteomics datasets, applies machine learning algorithms to prioritize PTM-specific amino acid substitutions and disrupted kinase-binding motifs, maps variants onto site-specific protein interaction networks and drug targets, and estimates the fraction of substitutions that affect PTM sites (~16–21%).