Fig. 1: Workflow of the analytic pipeline.
From: Polygenic contribution to the relationship of loneliness and social isolation with schizophrenia
GWAS summary statistics from schizophrenia83 and LNL-ISO32 were used. We evaluated the LNL-ISO polygenic score (PGSLNL-ISO) contribution to schizophrenia risk in an independent case-control sample (NSCZ = 1927; NHC = 1561). Subsequent genomic dissection of schizophrenia GWAS based on LNL-ISO led to different annotations: (i) SCZ[LNL]: variants from the schizophrenia GWAS associated with LNL-ISO (PLNL-ISO < 0.05), (ii) SCZ[CONC]: variants from the schizophrenia GWAS associated with LNL-ISO (PLNL-ISO < 0.05) and concordant allele effects in both phenotypes (BetaSCZ > 0 & BetaLNL-ISO > 0 OR BetaSCZ < 0 & BetaLNL-ISO < 0), and (iii) SCZ[DISC]: variants from the schizophrenia GWAS associated with LNL-ISO (PLNL-ISO < 0.05) and discordant allele effects in both phenotypes (BetaSCZ > 0 & BetaLNL-ISO < 0 OR BetaSCZ < 0 & BetaLNL-ISO > 0), and (iv) (SCZ[noLNL]: variants from the schizophrenia GWAS not associated with LNL-ISO (PLNL-ISO > 0.05); see Methods and Supplementary Methods for further details). We performed PGS analyses, partitioned heritability, and annotation-based stratified genetic covariance analyses across those subsets. We performed Mendelian randomization to evaluate causality between schizophrenia and LNL-ISO (and its constituent traits). “+” and “−” in the figure refer to the direction of the effect of the alleles studied.
