Title:	Bayesian inference benchmark in systems biology
DNr:	NAISS 2026/4-705
Project Type:	NAISS Small
Principal Investigator:	Henrik Häggström <henhagg@chalmers.se>
Affiliation:	Göteborgs universitet
Duration:	2026-04-13 – 2027-05-01
Classification:	10106
Keywords:

Abstract

In this project we perform a comprehensive benchmark of mainly Markov chain Monte Carlo (MCMC) methods for dynamical models in systems biology. The benchmark compares a number of state-of-the-art algorithms on both simulated data and real data scenarios. The scenarios covers a range of commonly encountered and challenging features such as multimodality, bifurcations, large scale problems and chaotic regimes. In addition, we develop a workflow package, based on the PEtab standard for parameter estimation problems, to allow users to easily perform reproducible and fair benchmarks. Main supervisor: Marija Cvijovic at the Department of Mathematical Sciences, Chalmers University of Technology and the University of Gothenburg.