Conjunctional false discovery rate

We used conjFDR implemented in the pleioFDR toolbox¹ to calculate the degree of shared genetic determinants between two traits. conjFDR is an extension of the condFDR method. condFDR and conjFDR leverage the pleiotropy of two associated traits to boost discovery of shared gene variants through an empirical Bayesian framework¹. condFDR boosts discovery of a trait of interest by supplementing it with genetic information from a second trait, which it factors in to the calculation of the test statistics with the aim of boosting discovery in the trait of interest. conjFDR expands on that approach by performing these test statistics adjustments for both of the traits in the comparison, providing insight into the shared genetic determinants between two traits².

In contrast to LD score regression, the samples of each of the two traits in a condFDR and conjFDR analysis must not overlap to avoid influence of the genomic inflation on the discovery. The procedure for both condFDR and conjFDR analyses start with calculating empirical cumulative distribution functions for the p-values of each SNP. It then applies an adjustments to the empirical CDFs p-values of the first trait based on the significance level of the p-values of the corresponding SNPs in the second trait. In conjFDR, it will apply the same process for the second trait conditioning on the significance levels of the first trait. The final conjFDR value (q_FDR) represents the maximum of the two adjusted p-values for each SNP. We used a threshold of q_FDR < 5e^-8 to determine whether a genetic locus was shared between two traits.

1.

Andreassen OA, Thompson WK, Schork AJ, Ripke S, Mattingsdal M, Kelsoe JR, et al. Improved Detection of Common Variants Associated with Schizophrenia and Bipolar Disorder Using Pleiotropy-Informed Conditional False Discovery Rate. PLOS Genetics [Internet]. 2013 Apr 25;9(4):e1003455. Available from: https://doi.org/10.1371/journal.pgen.1003455

2.

Smeland OB, Frei O, Shadrin A, O’Connell K, Fan CC, Bahrami S, et al. Discovery of shared genomic loci using the conditional false discovery rate approach. Human Genetics [Internet]. 2020 Jan 1;139(1):85–94. Available from: https://doi.org/10.1007/s00439-019-02060-2