Objective: To identify common and rare genetic variants which increase the risk of bipolar disorder (BP). Specific Objectives: 1. To discover both rare and common risk variants for BP by sequencing the whole genome of one affected individual from the 6 highest-density PIC families and the 36 highest-density families from the NIMH Genomics Initiative (42 patients in total), at ~32X coverage using second-generation short read DNA sequencers. 2. To rank the genes with greatest accumulations of deleterious variants discovered in SO 1 using a combination of bioinformatics criteria. We will implement an algorithm which prioritizes variants on the basis of functional changes and other features (e.g. exonic and promoter regions, and micro-RNA and transcription factor binding sites). This will be done in two steps: a. First, for variants discovered in regions linked to BP in that subject's family, and then b. In the rest of the genome. 3. Determine the complete DNA sequence of the 5 most promising genes in 500 BP cases and 500 normal controls by genome partitioning with Long PCR and second-generation short read DNA sequencers. 4. Impute novel variation into several large BP GWAS datasets. Currently, this only includes the Psychiatric GWAS Consortium (PGC, 7,481 cases and 9,250 controls). The Genomic Psychiatry Cohort and VA CSP#572 samples are also projected to be available. NB: This step will involve no new laboratory work or clinical assessments, but will only use existing GWAS marker data at that time. Background: Bipolar disorder is a major cause of disability amongst US veterans as well as worldwide. However, very little research in the genetics of BP has been done in the VA system. A number of genomewide association studies have reported several novel risk genes for BP but still explain only a small proportion of the genetic risk. Whole-genome sequencing has emerged in the last 2-3 years as the most comprehensive method to detecting genetic variation, and has recently resulted in several published findings of novel causes in several disorders. While prohibitively expensive only a few years ago, next-generation sequencing technologies have now made whole-genome sequencing possible, and it is currently being applied to complex diseases such as psychiatric illnesses. Proposed Methods: We plan to sequence the whole genomes of 42 patients with BP using the Illumina HiSeq 2000. To maximize the chance of identifying causative variants, these subjects will come from families in which there are at least 3 affected siblings. We will compare the genome sequences of these subjects to the sequence data in the 1000 Genomes Project, which is publicly available. We will prioritize sequence variants discovered based on their function. Based on our preliminary data, we expect to find thousands of deleterious variants which will not have been documented in established databases such as dbSNP. The 5 genes with the highest levels of deleterious variation will be sequenced in 500 cases and 500 controls using long PCR. Finally, we will attempt to impute the variants we discover in several large existing GWAS datasets, including one currently being collected in the VA system nationwide.
Bipolar disorder (BP) is a potentially devastating neuropsychiatric illness. There are an estimated 90,000 patients with BP in the VA system nationwide. There are currently no specific genetic sequence changes (mutations) known to cause illness across various ethnic groups, and a large proportion of patients do not respond to treatment adequately. Recently, we have developed the capacity to sequence the entire genomes of individuals. Finding potentially rare sequence changes that might cause the illness is a new and potentially powerful means to isolating the genetic risk factors for this illness. The identification of such mutations may provide vital clues to understanding not only the causes of the illness, but also to developing new treatments. In this project, we seek to determine the sequences of the entire genomes of individuals with BP and search for both common and rare mutations. We will then test the mutations that we discover in very large population- based samples.