SU25 - TARGETED NEXT GENERATION SEQUENCING OF 187 GENES IN A BULGARIAN PSYCHOSIS SAMPLE

Abstract

Background There is strong evidence for partial overlap of genetic influence on Schizophrenia (SCZ), Schizoaffective Disorder (SAD) and Bipolar Affective Disorder (BAD). Part of this is due to common genetic variants, detected by GWAS, with a small individual effect on risk. Recent studies using next generation sequencing reveal growing evidence of the role of rare genetic variants in both diseases. We performed targeted NGS on 187 candidate genes to further investigate the genetic architecture of Bulgarian patients with bipolar disorder and schizophrenia. Methods A total of 300 individuals with BAD, 151 with SCZ, 15 SAD according to DSMIV and ICD-10, as well as 85 healthy population controls and 40 healthy relatives were recruited. The samples were sequenced on the Ion PROTON platform. The sequencing panel comprised of 187 preselected strong candidate genes, based on the results from GWAS and previous NGS studies. Only samples with coverage of at least 95% of the target region at 20x were included in the analyses. Case-control association testing was done using PLINK. The rare variants have been evaluated and divided into groups based on their functional relevance: LOF variants (frameshift and nonsense); with potentially damaging effect on protein function (splice site, missense variants, 3’-UTR/5’-UTR variants); and with no known effect on function (synonymous and intron variants). Genes were prioritized by mutation burden analysis based on the average number of functional occurrences per 1000 bp of the gene CDS. The RVIS scoring system for assessing the intolerance of individual genes to protein-altering variants (Petrovski et al., 2013) was used in conjunction with the burden score for the final ordering. Results 5373 variants were detected, of which 2826 were found in affecteds (1831 singletons and 995 recurrent). No significant associations between cases and controls could be detected, that survived the correction for multiple testing. In total we found in patients 14 rare LOF variants; 889 non-synonymous variants (243 of which were potentially damaging); 109 splice site variants and 32 in regulatory regions. Discussion We could find no significant association of common or rare variants in the analyzed samples, most likely due to small sample size. However, we could identify both recurrent and singleton rare variants with potential functional relevance in genes associated with ion channels, postsynaptic plasticity, neurogenesis and signalling cascade pathways.