Informer

Informer prioritizes compounds for kinase inhibitor discovery by selecting informer compounds from chemogenomic datasets to predict activity against new kinase targets.

Key Features:

• Chemogenomic Data Utilization: Uses Published Kinase Inhibitor Sets (PKIS) bioactivity measurements as the source data for informer selection.
• Informer-Based-Ranking (IBR): Selects small sets of screening compounds ("informers") based on chemogenomic profiles to rank and predict other compounds' activities.
• Integration of Experimental Bioactivity: Combines chemogenomic information with experimentally measured bioactivity data in its analytical approach.
• Minimally-supervised Strategy: Applies a minimally-supervised computational strategy to transfer information from PKIS to new kinase targets.
• Prospective Testing and Validation: Validated using leave-one-out cross-validation and prospective testing, including evaluation on three kinase targets absent from PKIS.
• Application in Virtual Screening: Predicts compound activity to enrich screening libraries with likely actives against specified kinase targets.

Scientific Applications:

• Drug Discovery: Identifies and prioritizes candidate kinase inhibitors for early-stage therapeutic development.
• Compound Prioritization: Ranks untested compounds for experimental follow-up when direct target-specific data are limited.
• Virtual Screening: Enriches virtual screening outputs by predicting activity profiles against new kinase targets.

Methodology:

Employs a minimally-supervised approach integrating PKIS chemogenomic data with experimental bioactivity, selects informer compounds from PKIS to predict activities of untested compounds against new kinase targets, and validates predictions via leave-one-out cross-validation and prospective testing on novel kinase targets.