Matt Bratkowski
Associate Director Calico Life Sciences LLC
Matt Bratkowski is an Associate Director at Calico Life Sciences, where he leads the structural biology team. Prior to this, he was a Principal Scientist and Senior Scientist at Calico. He was previously a Senior Scientist and Scientist at Nura Bio (formerly Proneurotech). Matt has extensive expertise in structural biology, protein biochemistry, and drug discovery, with a career spanning both academia and industry. He completed postdoctoral research at UT Southwestern Medical Center working on X-ray crystallography of proteins related to gene transcription and earned his PhD in Biophysics from Cornell University working on RNA degradation
Seminars
- Benchmarking AI/ML models against relevant structural biology datasets to understand how performance varies across applications, targets, and data types
- Combining computational predictions with wet-lab validation to determine whether model outputs translate into experimentally meaningful improvements
- Establishing robust validation frameworks for AI adoption to give discovery teams greater confidence when selecting tools and investing them into SBDD workflows
AI and machine learning are rapidly becoming embedded within structural biology and drug discovery, but the research community is increasingly asking a more practical question: which approaches genuinely improve SBDD outcomes? With model performance varying by application, training data often limited, and automated structure solution and docking still facing significant challenges, integrating AI effectively requires rigorous validation and close collaboration with experimental structural biology. Join computational scientists, structural biologists, and AI developers as they discuss how to translate AI advances into measurable SBDD improvements by:
- Identifying where AI delivers genuine value across SBDD, evaluating applications beyond protein folding prediction, including structure prediction, cryo-EM processing, molecular design, docking, and experimental data analysis
- Validating AI/ML approaches against real-world experimental data, discussing how independent benchmarking, wet-lab validation, and appropriate performance measures can establish confidence in new models and distinguish useful tools from overpromoted solutions
- Integrating AI with experimental structural biology, exploring how prediction models and experimental methods can complement one another to improve speed, fidelity, resolution, and decision making across the SBDD workflow
