Servicenavigation

Project Details

Back

Projekt Print View

Read-Across the Targetome - An integrated structure- and ligand-based workbench for computational design of novel tool compounds

Subject Area Bioinformatics and Theoretical Biology
Term from 2017 to 2021
Project identifier Deutsche Forschungsgemeinschaft (DFG) - Project number 391684253
 
Final Report Year 2024

Final Report Abstract

The "Read-Across the Targetome" project aimed to develop a novel method to accelerate the discovery of chemical tool compounds for biomedical research. These compounds, known as "chemical probes," are essential for understanding protein functions in biological processes. However, suitable tool compounds are often unavailable for many important proteins, hindering research progress and therapeutic development. Existing initiatives like the Structural Genomics Consortium (SGC) have produced some chemical probes, but these cover only a small fraction of relevant proteins. The RATAR project aimed to address this gap by creating a comprehensive platform based on the "read-across" principle. The core idea is that proteins with similar binding sites are likely to interact with similar chemical compounds, allowing predictions for new chemical probes for previously unexplored proteins. The project involved integrating extensive data, including over 127,000 protein structures and 1.68 million chemical compounds from public databases like the Protein Data Bank (PDB) and ChEMBL, to systematically link proteins and their ligands to propose new chemical probes. The "read-across" approach, originally used in toxicology to predict chemical behavior based on similarity, was adapted here to identify structurally similar proteins. A new fingerprint based representation of protein binding sites was implemented. We initially focused on protein kinases, a target class involved in many diseases, to start with a more focussed approach due to their binding site similarity. This led to key methods like KiSSim, which compares kinase binding sites, generating kinome-wide trees to predict potential off-target effects. Additionally, KinFragLib, a library of novel kinase inhibitor fragments, was developed to support drug discovery. The proteome-wide method, the Ratar framework, was also further explored. It adapted existing encoding techniques to describe binding pockets, yet results suggested additional refinement will be needed for accurate pocket fragment matching. All project tools and datasets were released as open-source resources for public use.

Publications

 
 

Additional Information

© 2026 DFG Disclaimer / Copyright / Privacy Policy

Textvergrößerung und Kontrastanpassung