Incorporating gene pathway information in genetic association studies

Lead Research Organisation: London Sch of Hygiene and Trop Medicine
Department Name: Epidemiology and Population Health


The focus of much current research in the field of human genetics is on the search for genetic variation between individuals that may explain the variability in susceptibility to certain diseases observed in the population. The major difficulty encountered is the fact that the influence of genetic variants on diseases are very small. This project will investigate whether these signals can be reinforced by incorporating known information on how gene functions are related when searching for such variants.

Technical Summary

Genetic association studies on a large scale, including genomewide, are becoming increasingly common; however it is often difficult to disentangle truly associated variations from spurious ones, because of the vast number of hypotheses tested. Incorporation of prior biological knowledge, for instance regarding genetic pathways, may help in filtering out spurious associations and several methods (e.g. gene sets) have been successfully developed to address similar problems with gene expression data. The proposed project will investigate extensions of these methods to genetic association studies. As well as assessing the viability of simple methodologies based on sets and scan statistics, a modelling approach will also be considered whereby the selection of potentially associated genes is informed by prior knowledge on pathways likely to be involved in the disease at hand. The rationale is to give more weight to plausible genomic regions when searching for susceptibility variants and to boost association signals that appear to be consistent with a pathway relevant to the disease studied. This will also allow hypotheses to be investigated at the pathway level. Applications of these methods are envisaged using data from the British Heart Foundation 50k genomewide scan available through ongoing collaboration with the UCL genetics group.


10 25 50