Statistical design and inference for single cell gene expression data

The advances in biotechnology over the last few decades provide great opportunities for a better understanding of how our body works and how to keep it healthy and, through this, how diseases work and how to overcome them. These advances have allowed us to observe gene activity in tissue samples, and have revealed much about disease biology despite these samples typically being a mix of many different cell types. More recently, it has become possible to observe gene activity at the single cell level – overcoming problems concerning the heterogeneity of cells in tissue, but possibly introducing new issues.

Such data can reveal complex biological interactions between genes, describe the biological impact of external conditions, and identify the presence of complex and rare cell populations (e.g. in oncology and developmental biology). However, these data are often very noisy and biased by various technological constraints. Therefore, their statistical analysis is challenging, and a number of statistical questions remain unanswered. This project will attempt to address some of these gaps. In particular, we will be interested to compare data from tissue samples to single-cell experiments, and develop methodology for the joint statistical analysis of the two data types.
These will be motivated by a number of case studies such as a) understanding the effects of cell-signalling and environment on gene-expression, and b) understanding the heterogeneity of gene expression present in cancer.

An interest in molecular biology is required but previous experience is not necessary. Background in statistics and/or computational biology will be beneficial.