T13 - Gwas Of Bsnip Biofactors And Biotypes As Intermediate Phenotypes For Psychosis

Abstract

Background The Bipolar and Schizophrenia Network of Intermediate Phenotypes (BSNIP) is an ongoing NIMH-funded study aimed to identify biomarker-based classification of psychoses, with clinical, imaging and neurophysiological phenotypes, consistent with Research Domain Criteria. Analyses of the first BSNIP cohort (BSNIP1) reported 3 categorical Biotypes of patients derived from cluster analysis of 9 PCA-derived integrated measures of cognitive and neurophysiological experimental paradigms (Clementz, B. et al., Am J Psychiatry 173:4, April 2016). In the present work we ran Genome-Wide Association Studies using these Biotypes as phenotypes. Methods 711 Cases and 460 Control unrelated individuals from the BSNIP1 study were genotyped using the Illumina PsychChip followed by imputation using the 1000Genomes reference panel. 4,322,238 imputed SNP markers were included in GWAS using PLINK regression analysis, controlling for ethnic background with PCA. Analyses of each of the 3 categorical Biotypes vs. Control were performed, as well as each Biotype against the other two. Finally a GWAS was done using the quantitative scores derived from the paradigms' PCA analysis (“Biofactors”). Results After Bonferroni correction (α=1.15E-08), there were no significant associations. However we report here some suggestive and possible associations. Analysis of Biotype3 vs. Controls identified SNPs on the Chr.11p11.2 region with P=1.044E-07. GWAS of Biotype1 vs. Biotypes 2 & 3 found a marker on Chr.12q24.22 with P=2.411E-07. Results from quantitative trait analysis of Biofactors identified a SNP in the glutamate ionotropic receptor gene GRIN2B with the lowest P=9.50E-08. GRIN2B was also suggestively associated with the Sensorimotor Reactivity biofactor with a P=2.03E-07. Sensorimotor Reactivity also had a P=1.59E-07 association with the Chromosome 4 open reading frame 22 (C4orf22) located on Chr4q21.21. The Cognitive Control biofactor was suggestively associated with a locus on Chr.3p21.1 (P=1.52E-07). Discussion Despite the limited power of the sample size, some of our GWAS identified suggestively related SNPs to data-derived measures of experimental paradigms of psychosis (Biofactors) and distinguished one of the integrated clusters (Biotypes) from controls.