AI-Driven, Structure-Based Discovery of Bacterial Second Messenger Signaling Targets

Identifying potential binding protein targets for small molecules, such as substrates, transmitters, or drugs, is a major challenge in molecular biology and drug discovery alike. In biology, identifying the target protein is key for elucidating biochemical pathways and infection processes. In pharmacology, target identification for potential drug candidates derived from phenotypic screenings is indispensable for understanding the mechanism of action, exploring drug repurposing opportunities, and detecting potential side effects early on.

The substantial breakthroughs in computational structural biology enable the use of protein structure models alongside experimental structures (X-ray, cryo-EM) as the basis for a computational approach to target discovery. In this project, the doctoral candidate will apply and improve methods for structure-based target identification, tailoring them to specific target classes. To this end, classical molecular docking techniques and machine learning-driven approaches will be combined for predicting molecular complex structures. Based on heterogeneous data resources, new machine-learning models for target ranking will be developed. The resulting methods will be applied to a new class of anti-infective targets, with the opportunity for experimental validation.