Mega-analysis of brain structural covariance, genetics, and clinical phenotypes

Abstract

Normal and pathologic neurobiological processes influence brain morphology in coordinated ways that give rise to structural covariance patterns across brain regions and individuals. We present a mega-analysis of structural covariance with magnetic resonance imaging of 50,699 healthy and diseased individuals (12 studies, 130 sites, and 12 countries) over their lifespan (ages 5 through 97). Patterns of structural covariance (PSC) were highly heritable (0.05< h2 <0.78) and significantly associated with 1610 independent significant variants after Bonferroni correction (10.3 > -log10[p-value] > 8.8): 1245 previously unreported, and 69% of them independently replicated (-log10[p-value] = 4.5). Associations revealed an imaging phenotypic landscape between 2003 PSCs and 49 clinical and cognitive traits at multiple scales. We constructed machine learning-derived individualized imaging signatures for various disease diagnoses using PSC features at multiple scales, suggesting that disease effects on the brain were better manifested in a multi-scale continuum than on any single scale. Experimental results were integrated into the Multi-scale Structural Imaging Covariance (MuSIC) atlas and made publicly accessible through the BRIDGEPORT web portal (https://www.cbica.upenn.edu/bridgeport/). Our results reveal strong associations between brain structural covariance, genetics, and clinical phenotypes, supporting that PSCs can serve as an endophenotypic anatomic dictionary in future research.

Three Imaging Endophenotypes Characterize Neuroanatomical Heterogeneity of Autism Spectrum Disorder

“…The MR images of PHENOM and UK Biobank followed the same quality assurance procedure and were detailed elsewhere. 21,36…”

“…25 The genotyped and imputed SNPs (single nucleotide polymorphisms) from the UK Biobank were preprocessed with a standard quality check (QC) protocol (https://www.cbica.upenn.edu/bridgeport/data/pdf/BIGS_genetic_protocol.pdf). 36…”

“…Genetic correlations were estimated using LDSC 35 (unbiased from population overlaps). In particular, the GWAS summary statistics of ASD and SCZ dimensions were derived from the current study, and those of the other disease traits were from previous GWAS studies (see elsewhere 36 for details of the inclusion criteria.) *P<0.05 in the genetic correlation estimates.…”

“…25 The genotyped and imputed SNPs (single nucleotide polymorphisms) from the UK Biobank were preprocessed with a standard quality check (QC) protocol (https://www.cbica.upenn.edu/bridgeport/data/pdf/BIGS_genetic_protocol.pdf). 36 First, we extracted and excluded related individuals (2nd-degree related individuals) in the complete UK Biobank sample (N=488,377) via King software relationship inference. 61 Duplicated variants from all 22 autosomal chromosomes were then removed.…”

“…The details of the inclusion criteria were described elsewhere. 36 LDSC uses the GWAS summary statistics from our study and previous studies for the seven selected diseases to determine the proportion of variance that two clinical traits share due to genetic causes.…”