Principal Investigator: Dr Ines Barroso
Department: Human Genetics
Wellcome Trust Sanger Institute, Human Genetics, Wellcome Trust
Genome Campus, Cambridge CB10 1SA, United Kingdom
Tags: 14069, BMI, GWAS, obesity
1) Professor Audrey Hendicks
2) Professor Ismaa Farooqi
Collaborating Institutions and Addresses:
1) University of Colorado Denver, Mathematical and Statistical Sciences,
Academic Building 1 – 4217, 1201 Larimer Street, Denver, CO 80204,
2) University of Cambridge, Clinical Biochemistry, Metabolic Research Labs
Level 4, Institute of Metabolic Scinece Box 289, Addenbrooke’s Hospital,
Cambridge, CB2 0QQ, United Kingdom
Funding body: Internally funded through Wellcome Trust Sanger Institute core
1a: This proposal seeks access to UK Biobank data to support ongoing efforts to
identify genetic variants underlying human body weight by studying the
extremes of the BMI distribution (severe childhood obesity and lifelong
thinness). The UK Biobank data will be a valuable resource to validate findings in
our ongoing obesity studies and investigate the effect of identified loci in the
1b: Obesity is a major public health problem with substantial impacts on
morbidity and mortality. The research we plan is entirely congruent with the
stated aim of UK Biobank to improve “the prevention, diagnosis and treatment
of a wide range of serious and life-threatening illnesses”
1c: We are interested in how genes influence human body weight. In the full
data set of 500,000 participants we will generate cases and controls selected
from the top and bottom tails of the BMI distribution. We will then compare
these cases to each other and each to the “middle” of the distribution, to
identify correlations between genetic variation and extreme obesity/thinness.
These results will act as an independent replication set to validate any genetic
variants identified through our ongoing studies and will enable us to check the
effect of variants on BMI as a continous trait.
1d: We wish to study the full cohort.
Genetic effects can vary between individuals. For example, the effect of a genetic variant may depend on environmental exposures such as levels of physical activity, diet and other lifestyle covariates. However, currently available methods to analyse genotype-phenotype associations commonly ignore such interactions or consider interaction effects for a single discretised environmental exposure.
Explicitly modelling interactions may increase the power to detect trait-associated variants and additionally provide new insight of the functional mechanisms and the contexts in which such variants act. We have therefore developed a new statistical approach that explicitly accounts for heterogeneity in effect size and leverages structure in the data, achieved by defining a prior on effect sizes that encodes the similarity in environmental exposure between individuals.
We want to apply the method to the UK Biobank data to explore interaction effects between genotypes and available lifestyle factors, including but not limited to diet and physical activity levels, on adiposity measures including BMI and waist-hip ratio. We will additionally perform interaction analysis with existing GxE methods for comparative purposes.
Project extension – October 2019:
This proposal seeks access to UK Biobank data to support ongoing efforts to identify genetic variants underlying human body weight by studying the extremes of the BMI distribution (severe childhood obesity and lifelong thinness). The UK Biobank data will be a valuable resource to validate findings in our ongoing obesity studies and investigate the effect of identified loci in the general population. We also aim to study the genetic variants of extreme BMI (i.e. severe childhood obesity) by utilizing UK Biobank with other large sequencing datasets to which we have access. To ensure unbiased joint analysis of these datasets, we will develop novel methods to incorporate multiple sequencing datasets including external publicly available common control data such as the Genome Aggregation Database (gnomAD). We plan to study these methods across numerous diseases and traits to evaluate the performance of these methods across different disease characteristics such as genetic architecture, and disease prevalence.
Last updated Oct 15, 2019