Predicting changes in protein regulation and abundance in response to a perturbation is a fundamental task in analysis of complex biomolecular networks. However the task is far from trivial, and requires us to reason in presence of limited information and uncertainty. Fortunately, molecular biology is rich in structured causal knowledge that is qualitative in nature. For systems at steady state, we introduce an approach that represents the available information in form of a causal diagram, combines the diagram with observational steady-state measurements on network components, converts the information into a quantitative causal model and estimates of outcomes of interventions. The approach can produce accurate results even when only a subset of network components are experimentally observed. At the same time, despite the many advantages, such causal models struggle to model biomolecular systems with cyclic components, notably feedback loops and reversible reactions. Ordinary differential equation are often used in this case to infer the outcomes of interventions but require extensive time course measurements to estimate rate parameters. Therefore, for dynamical systems with cyclic components we introduce a transformation of biomolecular networks into causal models that captures the directed relationships between network components at steady state, eliminating the need for time course measurements, estimation of rate parameters, and computer simulations. We demonstrate the benefits of this approaches with real-life case studies.

Dr. Vitek is Raymond Bradford Bradstreet Professor in the Khoury College of Computer Sciences, and Director of the Barnett Institute for Chemical and Biological Analysis at Northeastern University. She holds a PhD in Statistics from Purdue University and was previously a Faculty and a University Faculty Scholar at Purdue.

Dr. Vitek's research intersects statistical science, machine learning, mass spectrometry, proteomics and systems biology. Statistical methods and open-source software developed in her lab, in particular MSstats for statistical analyses of quantitative proteomic experiments, and Cardinal for interpreting mass spectrometry imaging experiments, are widely used in academia and industry.

Dr. Vitek is an elected Fellow of the American Statistical Association, a recipient of the 2021 Gilbert S. Omenn Computational Proteomics Award of the US Human Proteome Organization (HUPO), and of the Indigo BioAutomation Females in Mass Spectrometry Distinguished Contribution Award. She is a recipient of the CAREER award of the National Science Foundation, and of the Essential Open-source Software Award of the Chan-Zuckerberg foundation.