Inferring the kinetic behaviour of regulatory networks from snapshot imaging single cell quantitative data.

Regulation of gene expression has been shown to be surprisingly dynamic and can be oscillatory, pulsatile or stochastically bursty in several instances. We have previously shown that the stochastic oscillatory behaviour of a transcription factor, Hes1, within the context of neural development controls cell-fate choices and their timing. However, it is not clear how general this oscillatory behaviour is, mainly because the generation of live-cell imaging data of transcription represents an experimental bottleneck.

In contrast to sparse live imaging data, experimenters are generating a large number of static single cell imaging quantitative data (known as `snapshot data’), such as the absolute mRNA and protein copy number, the location of the mRNA or the protein in the nucleus or the cytoplasm, transcription from one or two chromosomal loci, degradation rates etc. We propose to unlock the richness of of such snapshot data, to gain insight into the underlying molecular process. Therefore, there are two key aims in this project; one will be to apply Bayesian statistics methods in to infer kinetic information from static quantitative data obtained at several time points and from a large population of cells. Second, to develop computational models to predict the gene network structure underlying the gene expression dynamics. These predictions will then be tested by live imaging of transcription in a set of selected cases.

Training: The student will divide his/her time between the lab of Prof Nancy Papalopulu to obtain the single cell imaging data for this project, in the team of Prof. Magnus Rattray for Bayesian analysis, and the School of Physics and Astronomy in the team of Dr. Galla, where he/she will be building and analysing computational models of the genetic circuit. Day-to-day supervision will be provided by Research Associates in the Papalopulu lab, who are experimentalists with extensive supervisory experience. The theoretical part will be co-supervised by theoretical physicists (Galla group) and computational biologists (Rattray group) developing machine learning and Bayesian inference methods for single-cell data analysis. This supervisory team has worked with a shared student before, who successfully completed his thesis in 2016, with several joint publications, and is now a postdoctoral fellow in a computational/wet biology lab in Switzerland. The combination of supervisors will provide a unique opportunity for the student to acquire laboratory skills, experience in mathematical and computational modeling, and in the analysis of data. At the end of their PhD they will be an attractive candidate to work in the interface of Life and Physical Sciences.