UBC Theses and Dissertations
Differential RNA expression analysis across cancer clonal populations using Bayesian phylogenetic modelling Gillis, Sierra
Cancer cells follow evolutionary principles leading to clonal populations that are nearly genetically identical, called clones. Using single-cell DNA sequencing technologies, a copy number profile of each clone can be found. This copy number profile can then be used to infer the relatedness of the cells in the sample, represented by a phylogeny. Clades in the phylogeny are determined to be the clones based on similarity in mutation profiles. From the same tumour sample, single-cell RNA sequencing allows the measurement of the gene expression. These cells can then be matched to the clones based on the copy number changes and abundances in expression in those regions. From here, it is natural to compare the gene expression between the clones, as this phenotypic characterization can provide insight into treatment resistance and metastasis. Tools for differential expression analyses that compare the clones cannot determine exactly where in their evolutionary history the changes in gene expression occurred. We propose to use a Bayesian phylogenetic model to infer the expression states at ancestral, unobserved clones. This model takes as input the phylogeny determined from the DNA mutation profiles as well as RNA sequencing counts matched to the clones. The RNA sequencing data is binned into discrete values to be used in the discrete trait phylogenetic model. We use a Markov Chain Monte Carlo Algorithm to infer parameters of the model which are then used to infer the ancestral gene expression states. After the ancestral states have been inferred, we find genes that had changes in expression state along each branch. We test the performance of this approach on synthetic and real data, and demonstrate how the results can be used for downstream analysis to identify changes in expression during cancer evolution.
Item Citations and Data
Attribution-NonCommercial-NoDerivatives 4.0 International