T230. Polygenic Risk Score Increases Schizophrenia Liability Through Cognition-Relevant Pathways: Causal Modelling With Latent Cognition and Polygenic Risk Timothea Toulopoulou 1 , Xiaowei Zhang 2 , Stacey Cherny 2 , Dwight Dickinson 3 , Richard Straub 4 , Karen Berman 5 , Pak Sham 2 , and Daniel Weinberger 4 1 Bilkent University, 2 University of Hong Kong, 3 National Institutes of Health, 4 Lieber Institute for Brain Development, 5 National Institute of Mental Health Background: Cognition shares substantial genetic variance with schizophrenia, with recent evidence from GWAS data and from modeling of twin data suggesting direct causality from the former to the latter. However, it is not clear how much of the genetic component of schizophrenia is mediated through cognition. Thus, we included in causation models direct measurements of the genetic risk (e.g. schizophrenia polygenic risk) to quantify the genetic component of schizo- phrenia that is mediated by cognition and captured by the polygenic risk. Methods: Data were from 1,313 members of 1,078 families, and included 416 schizophrenia patients, 290 unaffected sib- lings, and 607 controls. Polygenic risk (PRS) were based on the latest data from the PGC and represented the sum of geno- typic scores for all common genetic variants associated with schizophrenia. Cognition (L-COG) was extracted through common pathway models and captured the common variance across measurements in six cognitive domains. Results: Of the genetic component of schizophrenia, 2.71% was through PRS pathways mediated by L-COG, 3.93% by PRS covariation pathways that included L-COG, and 26.87% by L-COG pathways not captured by the PRS. The remaining variance in schizophrenia liability was through pathways other than cognition and PRS. Conclusions: Cognition pathways captured by the PRS score mediated a significant part of genetic risk for schizophrenia. However, the evidence suggests that other cognition path- ways not captured by PRS mediate an even greater part. We anticipate that when schizophrenia PRS include all possible variants associated with risk, more than 25% of the variants’ cumulative effect will first influence variation in cognition. Supported By: X.Z. was funded by T.T. and P.S. through State Key Laboratory of Brain and Cognitive Sciences funds and an NIH subcontract (NIH-260850043) awarded to T.T. The data collection was funded by the intramural research program of the NIMH to the D.W.’s lab. Keywords: Polygenic Risk Score, Schizophrenia, Cognition, Data Modelling T231. Reduced Resting-State Functional Connectivity Between Cerebellar Lobules and Cortical Regions in Individuals With Schizophrenia Sarah V. Clark 1 , Vince D. Calhoun 2 , and Jessica A. Turner 1 1 Georgia State University, 2 The Mind Research Network & The University of New Mexico, Yale University, School of Medicine Background: Reduced resting-state functional connectivity between the cerebellum and cortex has been demonstrated in individuals with schizophrenia. However, most research has focused on specific regions, leaving out the broader context of cerebello-cortical connectivity. Therefore, this study investi- gated connectivity between individual lobules and the whole brain; we started with the left cerebellum to examine dys- functions in the right cortex in schizophrenia, which have been a point of contention. Relationships with polygenic risk scores (PGRS) were also explored. Methods: Eighty-five healthy controls and 73 individuals with schizophrenia underwent resting-state scans and genotyping. Imaging data were preprocessed using a standard pipeline. Resting-state functional connectivity was calculated between 10 left cerebellar lobules and the whole brain. Group differ- ences in connectivity between each lobule and the rest of the brain were computed. Relationships between PGRS (calcu- lated using PRSice) and connectivity were tested for main ef- fects and group x PGRS interactions. Results: Controls demonstrated stronger connectivity be- tween 7 cerebellar lobules and cortical regions. Significant connectivity differences were observed in the right parietal lobe, and to a lesser extent, frontal and temporal regions. PGRS did not predict cerebellar connectivity in either the control or schizophrenia group, but at the trend level higher PGRS score predicted lower cerebellar-cortical connectivity in several lobules. Conclusions: Reduced connectivity between cerebellar and cortical regions in the schizophrenia group supports the cognitive dysmetria theory, suggesting weaker regionally specific cerebellar-cortical connectivity in schizophrenia. Though the current dataset was likely too small to detect significant relationships, trends suggest greater genetic risk for schizophrenia may predict reduced cerebellar connectivity. Supported By: NIH/NIGMS P20GM103472; Georgia State University Neurogenomics 2CI Fellowship Keywords: Schizophrenia, Functional Connectivity, Cere- bellum, Polygenic Risk Score T232. Abnormal Thalamocortical Functional Connec- tivity Correlates With Sleep Spindle Deficits in Schizophrenia Bengi Baran 1 , Fikret Is ¸ ık Karahano glu 1 , Charmaine Demanuele 1 , Mark Vangel 1 , Robert Stickgold 2 , Alan Anticevic 3 , and Dara Manoach 1 1 Massachusetts General Hospital, 2 Beth Israel Deaconess Medical Center, 3 Yale University School of Medicine Background: A growing literature implicates abnormal thala- mocortical circuitry in the pathophysiology of schizophrenia. Resting state functional connectivity MRI (rs-fcMRI) studies consistently report increased thalamic connectivity with sensorimotor cortex both in schizophrenia patients and in in- dividuals at high risk, in whom it predicts conversion to psy- chosis. We examined the relation of thalamocortical functional connectivity to sleep spindles, defining EEG oscillations of non-rapid eye movement stage II sleep (N2) that are initiated in the thalamic reticular nucleus (TRN), depend on Poster Abstracts S218 Biological Psychiatry May 1, 2018; 83:S129eS455 www.sobp.org/journal Biological Psychiatry