| Popis: |
Variation in the mutation rate along the human genome, if not properly understood and accounted for, has the potential to confound evolutionary studies, produce spurious driver candidates in cancer studies, and hinder the diagnostics aimed at understanding the etiologies of genetic diseases. In this thesis I consider mutation rate variation in both the germ-line and somatic tissues, at varying scales. In the germ-line, the major advance of this thesis over previous works is use of direct methods to analyse the magnitude, scale and determinants of mutation rate variation. This has enabled us to tease apart the evolutionary and mutational forces, whilst directly quantifying the variation in the human mutation rate at different scales; at large scales the variation appears to be quite modest, however at the single nucleotide scale there is potentially huge cryptic variation in the mutation rate. I envisage that within the near future, the increase in de novo mutations coming from pedigree studies will allow for even greater understanding. I extend this work into somatic tissues, however due to the quality of data and heterogeneity of samples and cell types, the primary findings lean towards highlighting areas of improvement for next genereation sequencing (NGS) pipelines - I show ~4% of all single nucleotide variants from cancers appear to be errors - and develop methods with which future studies could provide more insight. With these methods and the ever increasing flow of somatic single nucleotide variants, coupled with the continual improvements in NGS technology, it should soon be possible to provide accurate answers to the questions posed of somatic mutation rate variation in this thesis. |
