Search

To increase power and minimize bias in statistical analyses, quantitative outcomes are often adjusted for precision and confounding variables using standard regression approaches. The outcome is modeled as a linear function of the precision variables and confounders; however, for many complex phenotypes, the assumptions of the linear regression models are not always met. As an alternative, we used neural networks for the modeling of complex phenotypes and covariate adjustments. We compared the prediction accuracy of the neural network models to that of classical approaches based on linear regression. Using data from the UK Biobank, COPDGene study, and Childhood Asthma Management Program (CAMP), we examined the features of neural networks in this context and compared them with traditional regression approaches for prediction of three outcomes: forced expiratory volume in one second (FEV1), age at smoking cessation, and log transformation of age at smoking cessation (due to age at smoking cessation being right-skewed). We used mean squared error to compare neural network and regression models, and found the models performed similarly unless the observed distribution of the phenotype was skewed, in which case the neural network had smaller mean squared error. Our results suggest neural network models have an advantage over standard regression approaches when the phenotypic distribution is skewed. However, when the distribution is not skewed, the approaches performed similarly. Our findings are relevant to studies that analyze phenotypes that are skewed by nature or where the phenotype of interest is skewed as a result of the ascertainment condition., Competing Interests: Michael H. Cho has received grant funding from GSK and Bayer, and consulting or speaking fees from AstraZeneca, Illumina, and Genentech. Edwin Silverman has received grant funding from Bayer and GSK. Ann Chen Wu has received grant funding from GSK. Dawn L. Demeo has received grant funding from Bayer and honoraria from Novartis.

While genome-wide association studies have identified genes involved in differential treatment responses to inhaled corticosteroids (ICS) in asthma, few studies have evaluated the potential effects of age in this context. A significant proportion of asthmatics experience exacerbations (hospitalizations and emergency department visits) during ICS treatment. We evaluated the interaction of genetic variation and age on ICS response (measured by the occurrence of exacerbations) through a genome-wide interaction study (GWIS) of 1,321 adult and child asthmatic patients of European ancestry. We identified 107 genome-wide suggestive (P<10-05) age-by-genotype interactions, two of which also met genome-wide significance (P<5x10-08) (rs34631960 [OR 2.3±1.6-3.3] in thrombospondin type 1 domain-containing protein 4 (THSD4) and rs2328386 [OR 0.5±0.3-0.7] in human immunodeficiency virus type I enhancer binding protein 2 (HIVEP2)) by joint analysis of GWIS results from discovery and replication populations. In addition to THSD4 and HIVEP2, age-by-genotype interactions also prioritized genes previously identified as asthma candidate genes, including DPP10, HDAC9, TBXAS1, FBXL7, and GSDMB/ORMDL3, as pharmacogenomic loci as well. This study is the first to link these genes to a pharmacogenetic trait for asthma., Competing Interests: The authors have declared that no competing interests exist.

Expression quantitative trait loci (eQTL) provide insight on transcription regulation and illuminate the molecular basis of phenotypic outcomes. High-throughput RNA sequencing (RNA-seq) is becoming a popular technique to measure gene expression abundance. Traditional eQTL mapping methods for microarray expression data often assume the expression data follow a normal distribution. As a result, for RNA-seq data, total read count measurements can be normalized by normal quantile transformation in order to fit the data using a linear regression. Other approaches model the total read counts using a negative binomial regression. While these methods work well for common variants (minor allele frequencies > 5% or 1%), an extension of existing methodology is needed to accommodate a collection of rare variants in RNA-seq data. Here, we examine 2 approaches that are direct applications of existing methodology and apply these approaches to RNAseq studies: 1) collapsing the rare variants in the region and using either negative binomial regression or Poisson regression and 2) using the normalized read counts with the Sequence Kernel Association Test (SKAT), the burden test for SKAT (SKAT-Burden), or an optimal combination of these two tests (SKAT-O). We evaluated these approaches via simulation studies under numerous scenarios and applied these approaches to the 1,000 Genomes Project., Competing Interests: The authors have declared that no competing interests exist.

Many well-powered genome-wide association studies have identified genetic determinants of self-reported smoking behaviors and measures of nicotine dependence, but most have not considered the role of structural variants, such as copy number variation (CNVs), influencing these phenotypes. Here, we included 2,889 African American and 6,187 non-Hispanic White subjects from the COPDGene cohort (http://www.copdgene.org) to carefully investigate the role of polymorphic CNVs across the genome on various measures of smoking behavior. We identified a CNV component (a hemizygous deletion) on chromosome 3p26.1 associated with two quantitative phenotypes related to smoking behavior among African Americans. This polymorphic hemizygous deletion is significantly associated with pack-years and cigarettes smoked per day among African American subjects in the COPDGene study. We sought evidence of replication in African Americans from the population based Atherosclerosis Risk in Communities (ARIC) study. While we observed similar CNV counts, the extent of exposure to cigarette smoking among ARIC subjects was quite different and the smaller sample size of heavy smokers in ARIC severely limited statistical power, so we were unable to replicate our findings from the COPDGene cohort. But meta-analyses of COPDGene and ARIC study subjects strengthened our association signal. However, a few linkage studies have reported suggestive linkage to the 3p26.1 region, and a few genome-wide association studies (GWAS) have reported markers in the gene (GRM7) nearest to this 3p26.1 area of polymorphic deletions are associated with measures of nicotine dependence among subjects of European ancestry., Competing Interests: Competing Interests: This study was supported by Award Number R01 HL089856, R01 HL089897 (EKS and JDC), K01HL125858 (SML), and R01 HL113264 (MHC) from the National Heart, Lung, and Blood Institute. The COPDGene® project is also supported by the COPD Foundation through contributions made to an Industry Advisory Board comprised of AstraZeneca, Boehringer Ingelheim, Novartis, Pfizer, Siemens, Sunovion, and GlaxoSmithKline. In the past three years, Edwin K. Silverman received honoraria and consulting fees from Merck, grant support and consulting fees from GlaxoSmithKline, and 331 honoraria and travel support from Novartis. The Atherosclerosis Risk in Communities Study is carried out as a collaborative study supported by National Heart, Lung, and Blood Institute contracts (HHSN268201100005C, HHSN268201100006C, HHSN268201100007C, HHSN268201100008C, HHSN268201100009C, HHSN268201100010C, HHSN268201100011C, and HHSN268201100012C), R01HL087641, R01HL59367 and R01HL086694; National Human Genome Research Institute contract U01HG004402; and National Institutes of Health contract HHSN268200625226C. Infrastructure was partly supported by Grant Number UL1RR025005, a component of the National Institutes of Health and NIH Roadmap for Medical Research. This support has not altered our adherence to PLOS ONE policies on sharing data and materials.