The interactive plot shows the genetic (below diagonal) and environmental (above diagonal) correlations for all non-binary phenotypes in the GeneATLAS. Hovering over a cell will provide information in a popup and the plot can be manipulated using the controls in the panel on the right.

A table of estimates of genetic parameters, including heritabilies as well as genetic, phenotypic and environmental covariances and correlations (however see the note at the end of the Methods below), for both non-binary and binary phenotypes is provided as a separate download [.csv.gz]  


In order to estimate heritabilities and genetic correlations we fitted a Linear Mixed Model by REML using the DISSECT software for each phenotype with a Genomic Relationship Matrix (GRM) containing common (MAF > 5%) autosomal genetic variants which passed QC (see manuscript). Using the model fits we computed best linear unbiased predictor estimates of genetic additive values for each individual. Using these the genetic correlations were estimated by computing correlations between the additive genetic values. Environmental correlations were estimated using the estimates of genetic correlations and estimates of phenotypic correlations, using the equation
$$ r_e = \frac{r_y – \sqrt{h^2_i h^2_j}r_g}{\sqrt{(1-h^2_i)(1-h^2_j)}} $$
where \(r_y,r_g\) are the estimates of phenotypic and genetic correlations and \(h^2_i,h^2_j\) are estimates of the heritability, on the observed scale, obtained from the variance component estimates of the LMM.
Due to the sparsity of cases for many binary phenotypes estimates of phenotypic correlations, and as a consequence environmental correlations, are very unstable. We therefore do not include any of these estimates in the provided download. If, despite these limitations, you are interested in an estimates for any pairs of the omitted binary phenotypes please do get in touch with us.