Deconvolution of bulk tumour transcriptome and DNA methylation data to reveal underlying cell populations.

The advent of high-throughput sequencing has boosted our understanding of the molecular mechanisms underlying cancer development and evolution to new levels. Often however, admixed normal cells hamper interpretation of the data deriving from such sequencing efforts. While single-cell sequencing can remedy these problems, it is experimentally involved and bulk tumour sequencing will likely remain the standard for the foreseeable future. In this project, we will hence develop methods to disentangle tumour bulk gene expression and DNA methylation data and reveal the distinct profiles of the normal and tumour cells. These methods will leverage as well as augment the wealth of cancer ‘omics data flowing from largescale consortia (e.g. the International Cancer Genome Consortium). Careful validation will come from teasing apart computationally mixed pure samples as well as from an extensive in-house single-cell sequencing project. To understand where further methodological advances may be made, we will carry out a detailed analysis of tumour expression and DNA methylation heterogeneity on these single-cell datasets. Finally, we will apply our methods in a pan-cancer setting. We believe that the deconvoluted expression/DNA methylation profiles will allow a more comprehensive taxonomy of cancer types. In turn this will lead to a better understanding of how genomic changes in the different cancers translate into changes in the transcriptome and epigenome to finally cause disease.