About this browser

The primary aim of the browser is to support interpretation of Genome Wide Association Studies (GWAS) results for Brain Imaging phenotypes and possible links to GWAS of other non-brain imaging phenotypes. The browser is built on top of data from several sources.

UK Biobank Brain Imaging Data

The primary source is results from 3,144 GWAS of Brain Imaging Derived Phenotypes (IDPs) measured on 9,707 participants of the UK Biobank study.

Association analysis was carried out using BGENIE and is described in full in this paper Elliott, L. et al (2018) https://www.nature.com/articles/s41586-018-0571-7

Summary statistics downloads for all 3,144 imaging GWAS will be available here shortly.

The genetic data and brain imaging data underlying this study are described in the following papers

Bycroft, C. et al (2018) https://www.nature.com/articles/s41586-018-0579-z

Miller, K. et al (2016) http://www.nature.com/neuro/journal/v19/n11/full/nn.4393.html

This project is a collaboration between

UK Biobank GWAS results

We have loaded the GWAS results of ~2,000 phenotypes in the UK Biobank processed by Ben Neale’s research group and available here

Other GWAS results

We have also included GWAS results from a number of other studies

Amyotrophic Lateral Sclerosis GWAS from van Rheenen et al. (2016). Genome-wide association analyses identify new risk variants and the genetic architecture of amytrophic lateral sclerosis. Nature Genetics 48, 9

Alzheimer’s disease GWAS from Jean-Charles Lambert et al. (2013) Meta-analysis of 74,046 individuals identifies 11 new susceptibility loci for Alzheimer's disease. Nature Genetics 45, 12

ADHD GWAS from the EAGLE consortium

Magnetic NMR GWAS from Kettunen et al. (2016) Genome-wide study for circulating metabolites identifies 62 loci and reveals novel systemic effects of LPA. Nature Communications 7, 11122

BMI, waist circumference and waist/hip ratio GWAS from the GIANT consortium

Allele coding, SNP effect sizes and p-values

For each SNP two alleles are reported with two alleles REF and ALT. So for example, at this SNP http://big.stats.ox.ac.uk:/variant/6:37161391-T-C the REF allele is T and the ALT allele is C. The browser also provides effect sizes and standard errors for each SNP tested against each phenotype. In all studies, the effect allele is the ALT allele. Consequently, the SNP effect sizes refer to the change in phenotype (quantitative traits) or log-odds (binary traits) corresponding to possession of an additional copy of the ALT allele.

A full interpretation, or comparison with other studies, also requires knowledge of any transformations performed on the phenotypes which are described in the references cited above. In particular, note that GWAS results for the UKBiobank phenotypes produced by the Neale lab are all based on transformed phenotypes.

In the PheWAS plots (for example http://big.stats.ox.ac.uk/variant/6-32605189-A-G) we have capped the p-values at –log10 p-value = 30 to enhance display and comparison. Exact p-values are available in the table below such plots.

Change log

17 Oct 2017 : Release of version 1 of the browser containing data from

Elliott, L. et al (2018) https://www.nature.com/articles/s41586-018-0571-7

06 Dec 2017 : Release of version 2 of the browser.

  • Addition of UK Biobank GWAS results from Neale group and other studies

  • Removal of Brain imaging functional connectivity IDPs relating to Netmat edges. This large number ~1,700 IDPs have low levels of heritability and show very few GWAS associations. Removing these IDPs improves the visualization of the remaining phenotypes. We have kept the ICA Netmat variables which do show significant non-zero heritability and several significant associations.

  • Effect sizes and standard errors are now included for each SNP.

  • Manhattan plots changed to have rainbow colour scheme.


For questions about this browser please email


This site was based on the PheWeb code base