Bioinformatics Seminars

Current Bioinformatics Seminar

26 September 2017

Statistical inference from cancer sequencing data

Simon Tavare
Cancer Research UK Cambridge Institute

Poolseq and cancer sequencing experiments produce aggregated data over different individuals or cells, for which conventional population genetics analysis methods do not seem appropriate. I will describe our ongoing attempts to understand what can be learned from site frequency data (e.g., the numbers of SNVs appearing in various proportions of cells of a tumour sample) obtained from such data. I will describe some theoretical aspects of modelling cancer evolution, in particular what seems to be easy to infer and what seems to be hard. The methodology comes from the ABC and coalescent part of the subject, with due allowance for the cancer setting. Time permitting, I will also outline our recently funded project from the CRUK Grand Challenge competition, which will provide novel in-situ data in 3.5D on a vast scale, along with some novel statistical problems.

Search past seminars