Inference of kinetics of large multi-scale chemical reaction networks

The main aim of the thesis is to fit mass spectroscopy data derived by chemists interested in the origin of life with the family of ”hierarchical models”. An adequate statistical method will be built which can be used to infer the parameters of the hierarchical models, which can be interpreted as ”proxy” kinetic rates. A Bayesian prior on the kinetic parameters k of potentially all mechanistically simple chemical reactions has been made available by recent work in computational chemistry, complementing chemical expertise, and allowing an inference of the network itself. Because only raw formulas are accessible through measurements, the general inference framework is that of HMM (hidden Markov models), for which a large panel of techniques have been developed, including expectation maximization (EM) and variational methods.

We are looking for a highly motivated mathematician or theoretical physicist with a background in statistical inference and/or statistical physics, and strong interest in applications and interference with scientists with very different backgrounds. Some proficiency in algorithmic programming in Python is required.