ELLIS–FIMM talk: Ola Spjuth "AI, Automation, and Phenomics for Mechanistic Insights: From Data Generation to Intelligent Agents"

AI, Automation, and Phenomics for Mechanistic Insights: From Data Generation to Intelligent Agents

Ola Spjuth
Uppsala University

Abstract: Understanding the mechanisms underlying cellular responses to chemical perturbations is a central challenge in biomedical research. In this talk, I will present how we combine automated phenomics platforms with scalable data generation and AI-based modeling to extract mechanistic insights from high-dimensional biological data. By leveraging high-content cellular profiling and machine learning, we move beyond descriptive assays toward models that offer insights into potential targets, modes of action, and cellular processes. I will highlight how automation and standardized experimental workflows enable the generation of large, high-quality datasets that support robust and reproducible AI modeling. These approaches allow us to systematically explore how compounds perturb cellular systems across conditions, doses, and time. Finally, I will discuss the emerging role of AI agents, autonomous systems capable of integrating heterogeneous data, generating hypotheses, and guiding experimental design, to accelerate scientific discovery. This points toward a future where biological insight generation becomes increasingly data-driven, iterative, and scalable.

This guest talk is hosted by Professor Samuel Kaski, Department of Computer Science.

Professor Spjuth will serve as the opponent for the PhD defence of Arslan Masood on the same day, at 12:00 in Hall T2. Find defence details here.