Biology subjects

Dahl, A.

Publications and source records attributed to Dahl, A..

58 records · Page 4Linked to original sources

bioRxiv2017

Effect of vitamin D supplementation on biomarkers of inflammation and immune function: functional genomics analysis of the BEST-D trial

Vitamin D deficiency has been associated with multiple diseases, but the causal relevance and underlying processes are not fully understood. Elucidating the mechanisms of action of drug treatments in humans is challenging, but application of functional genomic approaches in randomised trials may afford an opportunity to systematically assess molecular responses to treatments. In the Biochemical Efficacy and Safety Trial of Vitamin D (BEST-D), 305 community-dwelling individuals aged over 65 years were randomly allocated to treatment with vitamin D34000 IU, 2000 IU or placebo daily for 12 months. Genome-wide genotypes at baseline, and transcriptome and plasma levels of cytokines (IFN-{gamma}, IL-10, IL-8, IL-6 and TNF-) at baseline and after 12 months, were measured. The trial had >90% power to detect a 2-fold change in gene expression. Allocation to vitamin D for 12-months was associated with 2-fold higher plasma levels of 25-hydroxy-vitamin D (25[OH]D), but had no significant effect on whole-blood gene expression (FDR <5%) or on plasma levels of cytokines compared with placebo. In pre-specified analysis, rs7041 (intron variant, GC) had a significant effect on circulating levels of 25(OH)D in the low dose but not on the placebo or high dose vitamin D regimen. A gene expression quantitative trait locus analysis (eQTL) demonstrated evidence of 31,568 cis-eQTLs (unique SNP-probe pairs) among individuals at baseline and 34,254 after supplementation for 12 months (any dose), but had no significant effect on cis-eQTLs specific to vitamin D supplementation. The trial demonstrates the feasibility of application of functional genomics approaches in randomised trials to assess the effects of vitamin D on immune function.\n\nOne sentence summarySupplementation with high-dose vitamin D in older people for 12 months in a randomised, placebo-controlled trial had no significant effect on gene expression or on plasma concentrations of cytokines.\n\nTrial registrationSRCTN registry (Number 07034656) and the European Clinical Trials Database (EudraCT Number 2011-005763-24).\n\nFundingMedical Research Council, British Heart Foundation, Wellcome Trust, European Research Council and Clinical Trial Service Unit, Nuffield Department of Population Health, University of Oxford, Oxford, United Kingdom\n\nCopyrightOpen access article under the terms of CC BY.

clinical trials↗

bioRxiv2017

Singleton Variants Dominate the Genetic Architecture of Human Gene Expression

The vast majority of human mutations have minor allele frequencies (MAF) under 1%, with the plurality observed only once (i.e., \"singletons\"). While Mendelian diseases are predominantly caused by rare alleles, their cumulative contribution to complex phenotypes remains largely unknown. We develop and rigorously validate an approach to jointly estimate the contribution of all alleles, including singletons, to phenotypic variation. We apply our approach to transcriptional regulation, an intermediate between genetic variation and complex disease. Using whole genome DNA and lymphoblastoid cell line RNA sequencing data from 360 European individuals, we conservatively estimate that singletons contribute ~25% of cis-heritability across genes (dwarfing the contributions of other frequencies). Strikingly, the majority (~76%) of singleton heritability derives from ultra-rare variants absent from thousands of additional samples. We develop a novel inference procedure to demonstrate that our results are consistent with rampant purifying selection shaping the regulatory architecture of most human genes.

evolutionary biology↗

bioRxiv2017

Adjusting For Principal Components Of Molecular Phenotypes Induces Replicating False Positives

High-throughput measurements of molecular phenotypes provide an unprecedented opportunity to model cellular processes and their impact on disease. Such highly-structured data is strongly confounded, and principal components and their variants reliably estimate latent confounders. Conditioning on PCs in downstream analyses is known to improve power and reduce multiple-testing miscalibration and is an indispensable element of thousands of published functional genomic analyses. Further clarifying this approach is of fundamental interest to the genomics and statistics communities. We uncover a novel bias induced by PC conditioning and provide an analytic, deterministic and intuitive approximation. The bias exists because PCs are, roughly, unshielded colliders on a causal path: because PCs partially incorporate a causal genotype effect on one phenotype, the genotype becomes correlated with every phenotype conditional on PCs. We empirically quantify this bias in realistic simulations. For small genetic effects, a nearly negligible bias is observed for all tested PC variants. For large genetic effects, or other differential covariates, dramatic false positives can arise. Though one PC variant (supervised SVA) largely avoids this bias, it is computationally prohibitive genome-wide; further, its immunity to this bias is novel. Our analysis informs best practices for confounder correction in genomic studies.

genetics↗

bioRxiv2016

Statistical properties of simple random-effects models for genetic heritability

Random-effects models are a popular tool for analysing total narrow-sense heritability for simple quantitative phenotypes on the basis of large-scale SNP data. Recently, there have been disputes over the validity of conclusions that may be drawn from such analysis. We derive some of the fundamental statistical properties of heritability estimates arising from these models, showing that the bias will generally be small. We show that that the score function may be manipulated into a form that facilitates intelligible interpretations of the results. We use this score function to explore the behavior of the model when certain key assumptions of the model are not satisfied -- shared environment, measurement error, and genetic effects that are confined to a small subset of sites -- as well as to elucidate the meaning of negative heritability estimates that may arise.\n\nThe variance and bias depend crucially on the variance of certain functionals of the singular values of the genotype matrix. A useful baseline is the singular value distribution associated with genotypes that are completely independent -- that is, with no linkage and no relatedness -- for a given number of individuals and sites. We calculate the corresponding variance and bias for this setting.\n\nMSC 2010 subject classifications: Primary 92D10; secondary 62P10; 62F10; 60B20.

genetics↗