Petr Popov

Prof. Dr. Petr Popov is an Adjunct Professor of Applied Mathematics at the School of Computer Science & Engineering, Constructor University Bremen gGmbH. His research focuses on computational methods for molecular structures, biomolecular modeling, and drug discovery. He specializes in applying machine learning techniques to structural biology problems, including GPCR structure analysis and protein-ligand interaction prediction. Popov leads a lab developing innovative algorithms for binding site identification and thermostabilizing mutations in proteins. His work bridges bioinformatics and experimental structural biology, with applications in drug design and virology.

Research highlights include structural studies of cannabinoid receptors, serotonin receptors, and leukotriene receptors. He has pioneered graph neural network approaches for affinity prediction (graphDelta, graphLambda) and deep learning methods for druggable site detection. Popov collaborates with international teams on projects like CAPRI protein docking challenges and NIH-funded GPCR research. His lab's tools include PEPSI-Dock for protein-protein interactions and 3D convolutional networks for peptide binding site detection.

Popov's research has been published in top journals like Cell, Nature, and Science Advances. While no explicit awards are listed, his frequent collaborations with institutions like USC Bridge Institute and NIH indicate significant recognition in structural biology and computational drug discovery fields.

His current focus includes viral drug target identification using AI, molecular mechanisms of biased opioid signaling, and predicting disease-associated mutations in transmembrane proteins. The lab is also advancing machine learning models for high-accuracy binding affinity predictions and exploring fusion graph networks for drug discovery applications.