Computer-assisted drug design; Medicinal chemistry; Chemical Biology; Bioinformatics
Fuchs Jens-Alexander, Grisoni Francesca, Kossenjans Michael, Hiss Jan A., Schneider Gisbert (2018), Lipophilicity prediction of peptides and peptide derivatives by consensus machine learning, in
MedChemComm, 9(9), 1538-1546.
Schneider Gisbert (2017), Automating drug discovery, in
Nature Reviews Drug Discovery, 17(2), 97-113.
Gawehn Erik, Hiss Jan A., Schneider Gisbert (2016), Deep Learning in Drug Discovery, in
Molecular Informatics, 35(1), 3-14.
Reker D., Schneider P., Schneider G. (2016), Multi-objective active machine learning rapidly improves structure–activity models and reveals new protein–protein interaction inhibitors, in
Chemical Science, 7(6), 3919-3927.
Obtaining new chemical entities (NCEs) with a predictable macromolecular target engagement profile through rational molecular design is seminal for innovation in chemical biology and medicinal chemistry. G-protein coupled receptors (GPCRs) constitute an important class of macromolecular drug targets, for which the close intra-class relationships among GPCRs may easily result in undesired off-target modulation by druglike effector molecules. Under the research program described herein we aim at the multi-objective generation and computationally driven optimization of NCEs for cytokine receptors as promising anti-cancer targets in immunity and inflammation. We specifically target the C-C cytokine receptor 7 (CCR7) and C-X-C chemokine receptors 4 and 7 (CXCR4/7), while displaying negligible engagement of GPCR off-targets. The expected outcome will be first-in-class dual chemokine receptor blockers identified through computational molecular design and optimization. Advanced structure-based (receptor-based) pharmacophore models will aim at orthosteric and allosteric ligand binding sites, and ligand-based virtual screening and reaction-driven molecular de novo design will provide access to innovative chemotypes. By iterating through synthesize-test-learn cycles the computational structure-activity relationship model adapts to the given tasks. Batch and microfluidics-assisted synthesis as well as compound purchasing from a large pool of available screening compounds shall achieve the required experimental throughput. Direct receptor binding measurements will provide information about the binding affinity of the test compounds, and in vitro cell-based testing of the ligand effects on receptor signaling will complement the biophysical compound profiling. The most potent leads will be tested for computationally predicted off-target liabilities. Taking the challenging example of CCR7 and the twin CXCR4/7 receptor system, the study will result in new technology for rapidly generating GPCR-modulating lead molecules with designer polypharmacology and tool compounds for chemical biology and chemogenomics studies.