Choose timezone
Your profile timezone:
Seminars
Decision-making for workflows in computational biology, using AI and classical statistical approaches
by
→
Europe/London
R61 CR03 (RAL)
R61 CR03 (RAL)
Description
Modern computational science increasingly relies on complex workflows in which many decisions must be made: which data to process, which algorithms to apply, how their parameters should be chosen, and when a result is good enough to proceed. While similar challenges arise across many areas of computational science, they are important problems in computational biology, where workflows often involve many interdependent processing, modelling and analysis steps.
In this seminar, we will discuss early-stage ALC-funded projects at STFC exploring how we are addressing these problems within computational biology, using complementary approaches ranging from classical statistical methods to machine learning and AI. Information from previous computations can be used to understand the behaviour of scientific workflows, predict outcomes, and guide future decisions. Doing this reliably raises fundamental questions about how success should be defined and measured, how uncertainty and heterogeneous data should be handled, and how optimisation objectives can be aligned with the scientific outcomes we actually care about. Using examples from computational structural biology, we will introduce the problems we are tackling, the approaches we are developing, and some of the challenges involved in making better-informed decisions within complex computational workflows.
*****************************
Biographies of the speakers:
David McDonagh and Rob Nicholls are Computational Scientists in STFC Scientific Computing at the Rutherford Appleton Laboratory, working on methods, software and data analysis for computational structural biology.
David’s background is in crystal structure prediction, electronic structure theory, crystallography, software engineering, embedded devices and machine learning. He obtained a master’s in Natural Sciences at Leicester University, a master’s in Theory and Modelling in the Chemical Sciences at Oxford University, and a PhD in Theoretical and Computational Chemistry at Southampton University. His current work combines scientific software development with machine learning and AI to automate computational workflows.
Rob studied Mathematics at the University of York, followed by a master’s in Mathematics in the Living Environment and a PhD in Chemistry. He then spent over a decade at the MRC Laboratory of Molecular Biology in Cambridge, developing methods and software for macromolecular crystallography and cryo-electron microscopy. His research interests centre on statistical inference, modelling and data analysis, and he also teaches Applied Statistics at the University of Cambridge.