To simplify the sharing process, we reran the analyses from Karlsson Linnér et al. (2021), removing the 23andMe samples. To obtain the EXT-23andme summary statistics, we applied the EXT1.0 genomic structural equation model to the summary statistics of GWASs on externalizing behaviors and disorders that exclude 23andMe data. The EXT-23andMe model structure, fit, and parameters were similar to those of the original analysis; and the EXT1.0 and EXT-23andMe factors had similar genetic correlations with external traits. The variance in externalizing behaviors explained by the polygenic score for externalizing decreased from 8.9% to 8.4% for COGA and from 10.5% to 8.6% for Add Health when using the EXT-23andMe summary statistics to calculate polygenic scores. Those analyses are reported in Williams et al. (2023). To obtain these summary statistics (EXT–23andMe), please complete this form.
Note that all papers using these summary statistics must cite the original Externalizing GWAS manuscript and the manuscript describing the reduced model:
- Karlsson Linnér et al. Multivariate analysis of 1.5 million people identifies genetic associations with traits related to self-regulation and addiction. Nat. Neurosci. 24, 1367–1376 (2021).
- Williams et al. Guidelines for Evaluating the Comparability of Down-Sampled GWAS Summary Statistics. Behav. Genet (2023).