Natural selection influenced the genetic architecture of brain structure, behavioral and neuropsychiatric traits

Abstract

Natural selection has shaped the phenotypic characteristics of human populations.Genome-wide association studies (GWAS) have elucidated contributions of thousands of common variants with small effects on an individual's predisposition to complex traits (polygenicity), as well as wide-spread sharing of risk alleles across traits in the human phenome (pleiotropy). It remains unclear how the pervasive effects of natural selection influence polygenicity in brain-related traits. We investigate these effects by annotating the genome with measures of background (BGS) and positive selection, indications of Neanderthal introgression, measures of functional significance including loss-of-function (LoF) intolerant and genic regions, and genotype networks in 75 brain-related traits. Evidence of natural selection was determined using binary annotations of top 2%, 1%, and 0.5% of selection scores genome-wide.We detected enrichment (q<0.05) of SNP-heritability at loci with elevated BGS (7 phenotypes) and in genic (34 phenotypes) and LoF-intolerant regions (67 phenotypes). BGS (top 2%) significantly predicted effect size variance for trait-associated loci (σ 2 parameter) in 75 brainrelated traits (β=4.39x10 -5 , p=1.43x10 -5 , model r 2 =0.548). By including the number of DSM-5 diagnostic combinations per psychiatric disorder, we substantially improved model fit (σ 2 ~ BTop2% × Genic × diagnostic combinations; model r 2 =0.661). We show that GWAS with larger variance in risk locus effect sizes are collectively predicted by the effects of loci under strong BGS and in regulatory regions of the genome. We further show that diagnostic complexity exacerbates this relationship and perhaps dampens the ability to detect psychiatric risk loci.

Heterogeneity and Polygenicity in Psychiatric Disorders: A Genome-Wide Perspective

“…allele frequency differences due to unbalanced phenotype distribution). 48 The Research Domain Criteria (RDoC) paradigm compliments DSM and ICD classification systems by assessing clinical phenotypes hypothesized to more closely map onto underlying biological systems (e.g. neuroimaging data 49 and brain circuit activity 50 ).…”

Evidence of shared genetic influences underlying schizophrenia and alcohol use disorder, but not alcohol consumption

“…These differences are likely due to the co-occurrence of different mechanisms ( e . g ., evolutionary pressures and phenotypic heterogeneity) shaping the genetic architecture of psychiatric disorders and traits 18 .…”

“…Given recent evidence that AUD and SCZ have differing levels of polygenicity 17,18 , we used GENetic Effect-Size distribution Inference from Summary-level data 45 (GENESIS) to estimate the effect size distribution parameters for each disorder/trait (details given in Supplemental Material ). Effect size distributions are characterized by three descriptive statistics per GWAS: π c describes the proportion of susceptibility SNPs, σ 2 describes the variance in effect size for non-null SNPs, and α describes residual effects not captured by the variance of effect sizes such as population stratification, underestimated effects of extremely small-effect-size SNPs, and/or genomic deflation.…”

“…Because the extent to which the correlation between AUD and SCZ is attributable to salient functional categories remains unknown, we partitioned the genetic covariance between AUD and SCZ into relevant annotations. Finally, given recent evidence that AUD might be more polygenic than SCZ 17 , with a greater proportion of risk variants with small effect sizes 18 , we examined the genetic effect size distributions of AUD and SCZ. Across all approaches, we contrasted AUD with the largest GWAS of alcohol consumption 12 to assess whether the genetic overlap between SCZ and alcohol-related phenotypes is restricted to risk for pathological drinking or is also evident for a more widely assessed continuous measure of typical alcohol intake.…”

Genes identified in rodent studies of alcohol intake are enriched for heritability of human substance use

“…Alcohol consumption, alcohol use disorder (AUD), cigarette smoking and smoking cessation are all complex, heritable phenotypes, with twin and family estimates of heritability ranging from 20-70% 1,2 Molecular genetic studies have demonstrated that all of these behaviors are highly polygenic 3 , meaning that many common genetic variants of small effect sizes contribute to the variation in these phenotypes. Recent genome-wide association studies (GWASs) of these behaviors have leveraged collaborative efforts and increasing sample sizes (currently ranging from ∼40,000 to over a million individuals) to identify multiple genetic loci 4–7 .…”