DEVELOPMENT OF ANALYTICS FOR SINGLE CELL SEQUENCING OF CANCER GENOMES

Abstract

Single cell sequencing is promoted as a promising technique that can critically resolve the clonal structure within a tumour. However, the data generated is extremely noisy due to whole genome amplification. Therefore, it is necessary to understand the experimental noise introduced by whole genome amplification, so that bioinformatics tools can be tailored to overcome the noise. In this work, a comprehensive survey was first conducted to understand the properties of different amplification method. To this end, a bioinformatics workflow, SoVaTSiC (Somatic Variant Tool for Single Cell) was introduced. By applying SoVaTSiC to multiple cancer datasets, the results illustrate the possibility to characterise tumour heterogeneity and infer phylogenetic lineage, both of which will likely be useful for informing the treatment regime for a cancer patient. However, the results also suggest the need to integrate multiple omics datasets together, so that the pathways dysregulated in each tumour cell population can be elucidated.