Understanding the mechanisms of cellular processes is a major challenge in systems biology. Our current understanding relies on increasingly more sophisticated experimental techniques that, for instance, allow us to profile cellular identities in unprecedented detail. To fully exploit the wealth of data produced, we need to develop powerful new mathematical methods adapted specifically to these data sources and the biological problems they address.
Current analysis methods typically either: (1) reduce the dimension of the data so that it can be visualized (and therefore better comprehended) or (2) cluster the data (using supervised or unsupervised methods) in order to identify patterns that are representative of different biological conditions. The first of these techniques suffers from the limitation that projection to a lower dimensional space inevitably results in the loss of information; and the second presumes that clusters (which typically result from static gene expression patterns) are the only feature of interest. However, gene expression in the cell is not static, and many important genes are known to vary dynamically in their expression within individual cells.
Identifying de novo genetic oscillators has been extraordinarily difficult because these delicate dynamics are usually obscured when bulk measurements from large ensembles of cells are taken (which effectively just give an average of the dynamic state of all cells in the population).
This project therefore seeks to combine new high-throughout single-cell profiling with advances in data analytics and network modelling that are able to go beyond clustering to identify combinations of genes that oscillate in a coordinated way.
Principal Investigator: Dr Ruben Sanchez Garcia