Study of genes/SNPs association with the brain morphology, including in particular the sulci and the sulcal pits
Principal Investigator: Mr Vincent Frouin
Approved Research ID: 25251
Approval date: January 9th 2017
We propose to investigate the existence of genes or gene networks involved in brain cortical structures normal variability. In preliminary works, we have estimated the heritability of these cortical structures by studying a variety of features such as the depth of sulci. To compute these estimate we used separately a pedigree study with SOLAR (Sequential Oligogenic Linkage Analysis Routines) and an independent population with the genetic information using GCTA (Genome-wide Complex Trait Analysis). Our results suggest that there are regions for which brain cortical structures are more heritable. Thus, they are pertinent phenotypes to perform variant and gene association studies. Our research meets the UK Biobank's stated purpose because characterizing the genes that shape the cortical structures would help to build a reference set of ?non-pathological variability?-causing genes. This will in turn help to better identify and understand the potential genetic patterns underlying diseases for which the particular shapes of cortical structures have been shown to be a biomarker, such as polymicrogyria, schizophrenia or autism. We will extract the features of interest on structural images (MRI images). For this task, we will use the same pipelines we used to estimate the heritability; they include functions available in Freesurfer and Brainvisa (brainvisa.info) software. Then, we will perform SNPs (Single Nucleotide Polymorphism) association studies with the previously extracted features to identify the SNPs , We will test various genetic model and follow GWAS (genome-wide association study) good practices analysis using genetic tools like PLINK (whole genome association analysis toolset). Ultimately the IMAGEN cohort with similar content accessible from our lab will be used as a replicate. The statistical power of our study will be increased if we can include all the subjects with structural images and genetic data available in UK Biobank (currently around ~10000 subjects). This would provide us with better statistical power to characterize the largest possible number of SNPs associated with these cortical structures considered as phenotypes. Ideally, we would include subjects without known brain diseases in order not to bias our analyses. We detail the rational for the subject sample size in the ?expected value of any results? part.