Mutagenesis and its Biomedical Impact

Of the many thousands of DNA differences between individuals, only a minority have important contributions to disease risk or other traits that differ between them.
Finding those rare consequential differences is the basis for genetic diagnosis and predicting traits like height and drug response. It is also often the first step in understanding the basis of a disease at a molecular level.
As a community we can only efficiently find the consequential differences if they directly alter an encoded protein, the vast majority don’t. We are developing ways to understand the consequences of the majority of DNA changes regardless of whether they alter a protein.
Our approaches are often based on studying how large numbers of DNA sequence differences are inherited through generations of the human population. It allows us to tease apart two patterns, the pattern of new mutations and the pattern shaped by the health of people that carried those mutations.
The pattern of new mutations tells us about the underlying biology of the DNA, how it is replicated and repaired as well as showing how likely a piece of DNA is likely to be disrupted by a new mutation. The second pattern is what tells us if a type of DNA change is likely to have a consequence for human health. Although described here in the context of inherited DNA differences, we apply similar approaches to interpret the new mutations that arise in and drive the development of cancer.

Discriminating mutations of medical and functional importance from the many more that are of negligible biological consequence is a major unmet challenge for genomic medicine. This problem is particularly acute for sequence changes in the non-protein-coding majority of the genome. Solving this is a key component of realizing the MRC’s objective “to use genetics, imaging and biological indicators to understand predispositions to disease”. We will help address this problem by disentangling the mutually confounding patterns of mutation and selection, both of which if separately resolved can give insight into the biology of the genome and the phenotypic consequences of a DNA sequence change. Our three main aims are to: 1. Reveal the fine-scale mutation landscape of the genome and identify the processes that shape it. 2. Understand the molecular consequences of mutation at regulatory sites in the non-protein-coding genome. 3. Relate molecular phenotypes to organism biology through improved measures of selection that are applicable to the non-protein-coding genome.

Our approach is predominately computational, applying statistical analyses to somatic (cancer), population and between species variation obtained from genome sequencing studies. These data are intersected with transcription and chromatin state measurements and key missing data generated by collaborators such as the international FANTOM consortium, or ourselves in which case may involve obtaining and processing primary human tissue samples. Specific hypotheses generated from our initial analyses, such as mutagenic processes are explored experimentally using yeast or in vitro systems where tractable.