Multiple lines of evidence will improve power to identify genes involved in complex traits

Genetic variants associated with complex disorders have been identified using genome-wide association studies. However, the specific variant that has an impact on the trait often remains elusive. In-silico methods can help prioritise which variants should be used in functional studies.

I will use two novel methods to prioritise variant selection. I will explore published literature using Google’s SyntaxNet which overcomes limitations such as non-random selection of manually curated literature and unreliability of data in machine-curated literature. I will identify DNA sequence motifs that differ in frequency between regions with and without genetic risk variants and use such motifs to prioritise variants for follow up. I will publish power studies of both of these approaches using real data and create a tool that uses these two data sources to identify variants likely to be involved in complex disease.

This research will lead to the Integration of functional evidence based on previous work with text-based evidence and sequence motifs, the integration of pathway-based analysis and the application of the variant prioritisation technique to a large number of datasets.