The proposed research has one clear overarching objective. This is the application of causal analysis approaches to the investigation of the human gut microbiome. Proposed work will (i) generate predictors of the microbiome to assess causal contributions to disease, (ii) use existing causal predictors of human disease risk factors to assess the impact of these on the microbiome and (iii) use genetic variation previously associated with extremes of gut health to design a human Recall by Genotype (RbG) experiment which will assess the causal link between genetic risk, gut flora and pre-morbid disease.

The human gut microbiome is the collection of non-human organisms which abide and interact with the human gut. Although the faecal microbiome remains understudied as a proxy for gut flora, the beginnings of large, population based, collections of microbial variation are now showing important patterns suggestive of contributions to human health. Clinical and questionnaire-based covariates have not only been shown to be associated with microbiota composition, but they stably explain substantive portions of variation in gut flora. These first steps in the characterization and analysis of the microbiome at scale mark a new stage in the epidemiological analysis of the microbiome.

In parallel to this developing field, I have been developing and employing methods for causal analysis in complex epidemiological scenarios. My work has been intimately involved in the discovery of genetic associations with complex traits and the use of these in applied epidemiological contexts. Research has built on the development and application of causal analysis approaches and coincident work in the field of complex disease/phenotype genetics. The human microbiome is a logical target for the application of causal methods as (i) the microbiome field is attracting marked attention, but in a manner not directed by causal analyses, (ii) observational associations found with human microbiome measurement are strong and compelling, but are likely to be suffering from classical problems of observational epidemiology and (iii) if associations are causal, the human microbiome is an extremely attractive therapeutic target for the development of future therapeutic interventions as it is malleable. This proposal which will unite large-scale, population based, collections of human microbiome data and applied epidemiological analyses to allow causal analysis of relationships of variation in human gut flora.

This work will be challenging given complexities in measurement of the human gut microbiome, the assessment of genuine human autosomal genetic contributions to variation in patterns of gut flora, the application of results to the analysis of human disease and the ability to assess risk factors which themselves shape the microbiome. However, these challenges will be met by bringing together resources and expertise from the MRC Integrative Epidemiology Unit (IEU) at the University of Bristol (applied genetic epidemiology, PI Timpson and the Avon Longitudinal Study of Parents and Children), Leuven University (the human gut microbiome, collaborator Raes and the Flemish Gut Flora Project) and King’s College London (genetic epidemiology and the collection of human gut flora data, collaborator Spector and Twins UK).