Benjamin Maier: From Multi-Omics Data to Executable Mechanistic Models of Cell Signalling

North West Seminar Series of Mathematical Biology and Data Science University of Manchester (Host: William Morgans) Title: From Multi-Omics Data to Executable Mechanistic Models of Cell Signalling Speaker: Benjamin Maier (EMBL - European Bioinformatics Institute) Date: Mon 8th June 2026 Abstract: Cellular signalling allows organisms to communicate and adapt to environmental changes with deregulation leading to many diseases. Yet our current view, often summarized by annotated consensus pathways, fails to capture context-specific signal transduction rewiring and is biased towards well-studied processes. Multi-omics technologies now allow hypothesis-free, data-driven exploration of biological systems, phenotypes, perturbations, and disease states. However, the scale and noise of these data, together with the need for manual curation, make it difficult to systematically generate mathematical models. In this seminar, I will present a modular workflow for deriving context-specific, executable mathematical signalling models from multi-omics and phenotype data. The workflow builds on tools previously developed in the Petsalaki group: SELPHI2.0 and PhosX for kinase-substrate relationships and kinase-activity inference, phuEGO for active signalling module reconstruction from phosphoproteomics (v1) as well as spatial/single-cell transcriptomics data (v2), and CHARLIE for phenotype-specific network extraction. A key bottleneck in converting data-driven signalling networks into mathematical models is the lack of edge directionality and regulatory sign. FlowSign addresses this by predicting edge direction and regulation from prior knowledge, omics data, user annotations, and network topology. The resulting networks can be transformed into Boolean or ODE-based models to simulate context-specific signalling dynamics, test perturbation effects, and prioritize experimentally testable hypotheses. I will demonstrate this pipeline in ongoing projects on melanoma drug resistance, cardiac organoid design, and digital twins for rare-disease patients. Bio: Benji is a final-year Systems Biology PhD candidate in Evangelia Petsalaki’s Whole Cell Signalling group at EMBL-EBI and the University of Cambridge, as part of the EMBL International PhD Programme. His research focuses on data-driven and mechanistic modelling of cell signalling, including multi-omics network reconstruction, prediction of regulatory interaction signs, and building executable models for cardiac organoid design, melanoma drug resistance and rare-disease digital twins. He holds a B.Sc. in Biological Sciences from Heidelberg University and completed a joint M.Sc. programme in Molecular Techniques in Life Science at SciLifeLab (KTH Royal Institute of Technology, Stockholm University & Karolinska Institutet) in Stockholm. Petsalaki Group: Evangelia Petsalaki’s research group studies human cell signalling in healthy and disease conditions. The group uses interdisciplinary approaches, including data-driven network inference, modelling of cell processes and data integration, to understand how different environmental or genetic conditions affect cell signalling responses leading to diverse cell phenotypes. Their long-term aim is to create whole-cell signalling models, to better understand cell functions and disease. Evangelia moved to GSK in April 2026 and retains a 20% position at EMBL-EBI. Selected publications / Useful Reading: SELPHI2: Data-Driven Extraction of Human Kinase-Substrate Relationships From Omics Datasets (https://doi.org/10.1016/j.mcpro.2025....) PhosX: data-driven kinase activity inference from phosphoproteomics experiments (https://doi.org/10.1093/bioinformatic...) phuEGO: A Network-Based Method to Reconstruct Active Signaling Pathways From Phosphoproteomics Datasets (https://doi.org/10.1016/j.mcpro.2024....) Identification of phenotype-specific networks from paired gene expression-cell shape imaging data (https://doi.org/10.1101/gr.276059.121) Identification of phenotype-specific networks from paired gene expression-cell shape imaging data (https://doi.org/10.1038/s44320-025-00...) Decoding MASLD Progression: A Molecular Trajectory-Based Framework for Modelling Disease Dynamics (https://doi.org/10.1101/2025.01.14.63...)