Search

We consider the problem of variable screening in ultra-high-dimensional generalized linear models (GLMs) of nonpolynomial orders. Since the popular SIS approach is extremely unstable in the presence of contamination and noise, we discuss a new robust screening procedure based on the minimum density power divergence estimator (MDPDE) of the marginal regression coefficients. Our proposed screening procedure performs well under pure and contaminated data scenarios. We provide a theoretical motivation for the use of marginal MDPDEs for variable screening from both population as well as sample aspects; in particular, we prove that the marginal MDPDEs are uniformly consistent leading to the sure screening property of our proposed algorithm. Finally, we propose an appropriate MDPDE-based extension for robust conditional screening in GLMs along with the derivation of its sure screening property. Our proposed methods are illustrated through extensive numerical studies along with an interesting real data application.

BackgroundCancer genomic studies often include data collected from several omics platforms. Each omics data source contributes to the understanding of the underlying biological process via source specific (“individual”) patterns of variability. At the same time, statistical associations and potential interactions among the different data sources can reveal signals from common biological processes that might not be identified by single source analyses. These common patterns of variability are referred to as “shared” or “joint”. In this work, we show how the use of joint and individual components can lead to better predictive models, and to a deeper understanding of the biological process at hand. We identify joint and individual contributions of DNA methylation, miRNA and mRNA expression collected from blood samples in a lung cancer case–control study nested within the Norwegian Women and Cancer (NOWAC) cohort study, and we use such components to build prediction models for case–control and metastatic status. To assess the quality of predictions, we compare models based on simultaneous, integrative analysis of multi-source omics data to a standard non-integrative analysis of each single omics dataset, and to penalized regression models. Additionally, we apply the proposed approach to a breast cancer dataset from The Cancer Genome Atlas.ResultsOur results show how an integrative analysis that preserves both components of variation is more appropriate than standard multi-omics analyses that are not based on such a distinction. Both joint and individual components are shown to contribute to a better quality of model predictions, and facilitate the interpretation of the underlying biological processes in lung cancer development.ConclusionsIn the presence of multiple omics data sources, we recommend the use of data integration techniques that preserve the joint and individual components across the omics sources. We show how the inclusion of such components increases the quality of model predictions of clinical outcomes.

Background: Cancer genomic studies often include data collected from several omics platforms. Each omics data source contributes to the understanding of the underlying biological process via source specific (”individual”) patterns of variability. At the same time, statistical associations and potential interactions among the different data sources can reveal signals from common biological processes that might not be identified by single source analyses. These common patterns of variability are referred to as ”shared” or ”joint”. To capture both contributions of variance, integrative dimension reduction techniques are needed. Integrated PCA is a model based generalization of principal components analysis that separates shared and source specific variance by iteratively estimating covariance structures from a matrix normal distribution. Angle based JIVE is a matrix factorization method that decomposes joint and individual variation by permutation of row subspaces. We apply these techniques to identify joint and individual contributions of DNA methylation, miRNA and mRNA expression collected from blood samples in a lung cancer case control study nested within the Norwegian Woman and Cancer (NOWAC) cohort study.Results: In this work, we show how an integrative analysis that preserves both components of variation is more appropriate than analyses considering uniquely individual or joint components. Our results show how both joint and individual components contribute to a better quality of model predictions, and facilitate the interpretation of the underlying biological processes.Conclusions: When compared to a non integrative analysis of the three omics sources, integrative models that simultaneously include joint and individual components result in better prediction of cancer status and metastatic cancer at diagnosis.

Quantitative genetic analyses require extensive measurements of phenotypic traits, a task that is often not trivial, especially in wild populations. On top of instrumental measurement error, some traits may undergo transient (i.e., nonpersistent) fluctuations that are biologically irrelevant for selection processes. These two sources of variability, which we denote here as measurement error in a broad sense, are possible causes for bias in the estimation of quantitative genetic parameters. We illustrate how in a continuous trait transient effects with a classical measurement error structure may bias estimates of heritability, selection gradients, and the predicted response to selection. We propose strategies to obtain unbiased estimates with the help of repeated measurements taken at an appropriate temporal scale. However, the fact that in quantitative genetic analyses repeated measurements are also used to isolate permanent environmental instead of transient effects requires that the information content of repeated measurements is carefully assessed. To this end, we propose to distinguish “short‐term” from “long‐term” repeats, where the former capture transient variability and the latter help isolate permanent effects. We show how the inclusion of the corresponding variance components in quantitative genetic models yields unbiased estimates of all quantities of interest, and we illustrate the application of the method to data from a Swiss snow vole population. This is the peer reviewed version of an article, which has been published in final form at [https://doi.org/10.1111/evo.13573]. This article may be used for non-commercial purposes in accordance with Wiley Terms and Conditions for Self-Archiving.

Studies on the effects of air pollution and more generally environmental exposures on health require measurements of pollutants, which are affected by measurement error. This is a cause of bias in the estimation of parameters relevant to the study and can lead to inaccurate conclusions when evaluating associations among pollutants, disease risk and biomarkers. Although the presence of measurement error in such studies has been recognized as a potential problem, it is rarely considered in applications and practical solutions are still lacking. In this work, we formulate Bayesian measurement error models and apply them to study the link between air pollution and omic signals. The data we use stem from the "Oxford Street II Study", a randomized crossover trial in which 60 volunteers walked for two hours in a traffic-free area (Hyde Park) and in a busy shopping street (Oxford Street) of London. Metabolomic measurements were made in each individual as well as air pollution measurements, in order to investigate the association between short-term exposure to traffic related air pollution and perturbation of metabolic pathways. We implemented error-corrected models in a classical framework and used the flexibility of Bayesian hierarchical models to account for dependencies among omic signals, as well as among different pollutants. Models were implemented using traditional Markov Chain Monte Carlo (MCMC) simulative methods as well as integrated Laplace approximation. The inclusion of a classical measurement error term resulted in variable estimates of the association between omic signals and traffic related air pollution measurements, where the direction of the bias was not predictable a priori. The models were successful in including and accounting for different correlation structures, both among omic signals and among different pollutant exposures. In general, more associations were identified when the correlation among omics and among pollutants were modeled, and their number increased when a measurement error term was additionally included in the multivariate models (particularly for the associations between metabolomics and NO2).

Exposure to traffic-related air pollution (TRAP) has been associated with adverse health outcomes but underlying biological mechanisms remain poorly understood. Two randomized crossover trials were used here, the Oxford Street II (London) and the TAPAS II (Barcelona) studies, where volunteers were allocated to high or low air pollution exposures. The two locations represent different exposure scenarios, with Oxford Street characterized by diesel vehicles and Barcelona by normal mixed urban traffic. Levels of five and four pollutants were measured, respectively, using personal exposure monitoring devices. Serum samples were used for metabolomic profiling. The association between TRAP and levels of each metabolic feature was assessed. All pollutant levels were significantly higher at the high pollution sites. 29 and 77 metabolic features were associated with at least one pollutant in the Oxford Street II and TAPAS II studies, respectively, which related to 17 and 30 metabolic compounds. Little overlap was observed across pollutants for metabolic features, suggesting that different pollutants may affect levels of different metabolic features. After observing the annotated compounds, the main pathway suggested in Oxford Street II in association with NO2 was the acyl-carnitine pathway, previously found to be associated with cardio-respiratory disease. No overlap was found between the metabolic features identified in the two studies., Highlights • Two randomized crossover trials were used to assess the relationship between TRAP and metabolic features with MS-based metabolomics (MWAS) • The locations represent different exposure scenarios, with London characterized by diesel vehicles and Barcelona by normal mixed urban traffic • Levels of 17 and 30 metabolic compounds associated with different air pollutants in the studies, with little overlap in features across pollutants • No overlap found between metabolomic features identified in the two studies, possibly due to different levels of single pollutants • The acyl-carnitine pathway, involved in cardio-respiratory disease, was suggested as a potential pathway in association with NO2 in one study

Measurement error and other forms of uncertainty are commonplace in ecology and evolution and may bias estimates of parameters of interest. Although a variety of approaches to obtain unbiased estimators are available, these often require that errors are explicitly modeled and that a latent model for the unobserved error-free variable can be specified, which in practice is often difficult.Here we propose to generalize a heuristic approach to correct for measurement error, denoted as simulation extrapolation (SIMEX), to situations where explicit error modeling fails. We illustrate the application of SIMEX using the example of estimates of quantitative genetic parameters, e. g. inbreeding depression and heritability, in the presence of pedigree errors. Following the original SIMEX idea, the error in the pedigree is progressively increased to determine how the estimated quantities are affected by error. The observed trend is then extrapolated back to a hypothetical error-free pedigree, yielding unbiased estimates of inbreeding depression and heritability. We term this application of the SIMEX idea to pedigrees “PSIMEX”. We tested the method with simulated pedigrees with different pedigree structures and initial error proportions, and with real field data from a free-living population of song sparrows.The simulation study indicates that the accuracy and precision of the extrapolated error-free estimate for inbreeding depression and heritability are good. In the application to the song sparrow data, the error-corrected results could be validated against the actual values thanks to the availability of both an error-prone and an error-free pedigree, and the results indicate that the PSIMEX estimator is close to the actual value. For easy accessibility of the method, we provide the novel R-package PSIMEX.By transferring the SIMEX philosophy to error in pedigrees, we have illustrated how this heuristic approach can be generalized to situations where explicit latent models for the unobserved variables or for the error of the variables of interest are difficult to formulate. Thanks to the simplicity of the idea, many other error problems in ecology and evolution might be amenable to SIMEX-like error correction methods.

DNA methylation changes are associated with cigarette smoking. We used the Illumina Infinium HumanMethylation450 array to determine whether methylation in DNA from pre-diagnostic, peripheral blood samples is associated with lung cancer risk. We used a case-control study nested within the EPIC-Italy cohort and a study within the MCCS cohort as discovery sets (a total of 552 case-control pairs). We validated the top signals in 429 case-control pairs from another 3 studies. We identified six CpGs for which hypomethylation was associated with lung cancer risk: cg05575921 in the AHRR gene (p-valuepooled = 4 × 10-17 ), cg03636183 in the F2RL3 gene (p-valuepooled = 2 × 10 - 13 ), cg21566642 and cg05951221 in 2q37.1 (p-valuepooled = 7 × 10-16 and 1 × 10-11 respectively), cg06126421 in 6p21.33 (p-valuepooled = 2 × 10-15 ) and cg23387569 in 12q14.1 (p-valuepooled = 5 × 10-7 ). For cg05951221 and cg23387569 the strength of association was virtually identical in never and current smokers. For all these CpGs except for cg23387569, the methylation levels were different across smoking categories in controls (p-valuesheterogeneity ≤ 1.8 x10 - 7 ), were lowest for current smokers and increased with time since quitting for former smokers. We observed a gain in discrimination between cases and controls measured by the area under the ROC curve of at least 8% (p-values ≥ 0.003) in former smokers by adding methylation at the 6 CpGs into risk prediction models including smoking status and number of pack-years. Our findings provide convincing evidence that smoking and possibly other factors lead to DNA methylation changes measurable in peripheral blood that may improve prediction of lung cancer risk.

Lifestyle factors, such as food choices and exposure to chemicals, can alter DNA methylation and lead to changes in gene activity. Two such exposures with pharmacologically active components are coffee and tea consumption. Both coffee and tea have been suggested to play an important role in modulating disease-risk in humans by suppressing tumour progression, decreasing inflammation and influencing estrogen metabolism. These mechanisms may be mediated by changes in DNA methylation. To investigate if DNA methylation in blood is associated with coffee and tea consumption, we performed a genome-wide DNA methylation study for coffee and tea consumption in four European cohorts (N = 3,096). DNA methylation was measured from whole blood at 421,695 CpG sites distributed throughout the genome and analysed in men and women both separately and together in each cohort. Meta-analyses of the results and additional regional-level analyses were performed. After adjusting for multiple testing, the meta-analysis revealed that two individual CpG-sites, mapping to DNAJC16 and TTC17, were differentially methylated in relation to tea consumption in women. No individual sites were associated with men or with the sex-combined analysis for tea or coffee. The regional analysis revealed that 28 regions were differentially methylated in relation to tea consumption in women. These regions contained genes known to interact with estradiol metabolism and cancer. No significant regions were found in the sex-combined and male-only analysis for either tea or coffee consumption.

DNA hypomethylation in certain genes is associated with tobacco exposure but it is unknown whether these methylation changes translate into increased lung cancer risk. In an epigenome-wide study of DNA from pre-diagnostic blood samples from 132 case–control pairs in the NOWAC cohort, we observe that the most significant associations with lung cancer risk are for cg05575921 in AHRR (OR for 1 s.d.=0.37, 95% CI: 0.31–0.54, P-value=3.3 × 10−11) and cg03636183 in F2RL3 (OR for 1 s.d.=0.40, 95% CI: 0.31–0.56, P-value=3.9 × 10−10), previously shown to be strongly hypomethylated in smokers. These associations remain significant after adjustment for smoking and are confirmed in additional 664 case–control pairs tightly matched for smoking from the MCCS, NSHDS and EPIC HD cohorts. The replication and mediation analyses suggest that residual confounding is unlikely to explain the observed associations and that hypomethylation of these CpG sites may mediate the effect of tobacco on lung cancer risk., Smoking tobacco is known to alter DNA methylation. Here, the authors show that hypomethylation of smoke-related genes is associated with future increase in lung cancer risk.

Low socioeconomic status (SES) is associated with earlier onset of age-related chronic conditions and reduced life-expectancy, but the underlying biomolecular mechanisms remain unclear. Evidence of DNA-methylation differences by SES suggests a possible association of SES with epigenetic age acceleration (AA). We investigated the association of SES with AA in more than 5,000 individuals belonging to three independent prospective cohorts from Italy, Australia, and Ireland. Low SES was associated with greater AA (β = 0.99 years; 95% CI 0.39,1.59; p = 0.002; comparing extreme categories). The results were consistent across different SES indicators. The associations were only partially modulated by the unhealthy lifestyle habits of individuals with lower SES. Individuals who experienced life-course SES improvement had intermediate AA compared to extreme SES categories, suggesting reversibility of the effect and supporting the relative importance of the early childhood social environment. Socioeconomic adversity is associated with accelerated epigenetic aging, implicating biomolecular mechanisms that may link SES to age-related diseases and longevity.