MTAG: Multi-Trait Analysis of GWAS implicates novel loci for depressive symptoms, neuroticism, and subjective well-being

The standard approach in genome-wide association studies (GWAS) is to meta-analyze association statistics from cohort-level GWAS of a single trait. Such single-trait analyses do not exploit information that may be available... [ view full abstract ]

The standard approach in genome-wide association studies (GWAS) is to meta-analyze association statistics from cohort-level GWAS of a single trait. Such single-trait analyses do not exploit information that may be available from GWAS of other correlated traits. We introduce a method, Multi-Trait Analysis of GWAS (MTAG), which jointly analyzes GWAS results for several related traits, thereby boosting statistical power to detect genetic associations for each trait¹.

MTAG is a generalization of standard, inverse-variance-weighted meta-analysis. The estimator takes summary statistics from at least two single-trait GWASs and, after jointly analyzing these, outputs trait-specific association statistics. The set of results for each trait represents an optimal combination of the information from the single-trait summary statistics, accounting for sample overlap and correlated genetic effects between the GWAS. The resulting p values can be interpreted like GWAS p values and used in the usual ways, for example, to prioritize SNPs for subsequent analyses including biological annotation.

We demonstrate MTAG using data on depressive symptoms (DEP; effective N = 354,862), neuroticism (NEUR; N = 168,105), and subjective well-being (SWB; N = 388,538). Comparing single-trait GWAS to MTAG, the number of genome-wide significant loci increases from 32 to 74 for DEP, from 9 to 66 for NEUR, and from 13 to 60 for SWB. These gains in statistical power from MTAG analyses are equivalent to increasing the sample size of the original, single-trait GWASs by 37% (DEP), 96% (NEUR), and 85% (SWB).

The association statistics from MTAG are replicated in a large, well-phenotyped cohort (N=6,857-8,307 for the three phenotypes) and improve prediction accuracy of polygenic risk scores by approximately 25%, consistent with theoretical calculations. Moreover, the MTAG results yield more informative bioinformatics analyses, suggesting a role of glutamatergic neurotransmission in depression.

References:

¹Turley, P. et al. MTAG: Multi-Trait Analysis of GWAS. bioRxiv (2017).