Penalized logistic regression for high-dimensional DNA methylation data with case-control studies

AbstractIn genetic epidemiology, rare variant case-control studies aim to investigate the association between rare genetic variants and human diseases. Rare genetic variants lead to sparse covariates that are predominately zeros and this sparseness leads to estimators of log-OR parameters that are biased away from their null value of zero. Different penalized-likelihood methods have been developed to mitigate this sparse-data bias for case-control studies. In this research article, we study penalized logistic regression using a class of log-F priors indexed by a shrinkage parameter m to shrink the biased MLE towards zero. We propose a maximum marginal likelihood method for estimating m, with the marginal likelihood obtained by integrating the latent log-ORs out of the joint distribution of the parameters and observed data. We consider two approximate approaches to maximizing the marginal likelihood: (i) a Monte Carlo EM algorithm and (ii) a combination of a Laplace approximation and derivative-free optimization of the marginal likelihood. We evaluate the statistical properties of the estimator through simulation studies and apply the methods to the analysis of genetic data from the Alzheimer’s Disease Neuroimaging Initiative (ADNI).

Download Full-text

Network-based regularization for matched case-control analysis of high-dimensional DNA methylation data

Statistics in Medicine ◽

10.1002/sim.5694 ◽

2012 ◽

Vol 32 (12) ◽

pp. 2127-2139 ◽

Cited By ~ 17

Author(s):

Hokeun Sun ◽

Shuang Wang

Keyword(s):

Dna Methylation ◽

Case Control ◽

High Dimensional ◽

Control Analysis ◽

Methylation Data ◽

Matched Case ◽

Case Control Analysis

Download Full-text

Power and sample size for multivariate logistic modeling of unmatched case-control studies

Statistical Methods in Medical Research ◽

10.1177/0962280217737157 ◽

2017 ◽

Vol 28 (3) ◽

pp. 822-834

Author(s):

Mitchell H Gail ◽

Sebastien Haneuse

Keyword(s):

Logistic Regression ◽

Sample Size ◽

Case Control ◽

Simulation Methods ◽

Case Control Studies ◽

Control Data ◽

Logistic Analysis ◽

Sample Size Calculations ◽

Control Designs ◽

Univariate Analyses

Sample size calculations are needed to design and assess the feasibility of case-control studies. Although such calculations are readily available for simple case-control designs and univariate analyses, there is limited theory and software for multivariate unconditional logistic analysis of case-control data. Here we outline the theory needed to detect scalar exposure effects or scalar interactions while controlling for other covariates in logistic regression. Both analytical and simulation methods are presented, together with links to the corresponding software.

Download Full-text

Association of FOXO3 Blood DNA Methylation with Cancer Risk, Cancer Survival, and Mortality

Cells ◽

10.3390/cells10123384 ◽

2021 ◽

Vol 10 (12) ◽

pp. 3384

Author(s):

Chenglong Yu ◽

Allison M. Hodge ◽

Ee Ming Wong ◽

Jihoon Eric Joo ◽

Enes Makalic ◽

...

Keyword(s):

Lung Cancer ◽

Dna Methylation ◽

Cancer Risk ◽

Cancer Survival ◽

Conditional Logistic Regression ◽

B Cell Lymphoma ◽

Case Control ◽

Case Control Studies ◽

Cox Models ◽

Total N

Genetic variants in FOXO3 are associated with longevity. Here, we assessed whether blood DNA methylation at FOXO3 was associated with cancer risk, survival, and mortality. We used data from eight prospective case–control studies of breast (n = 409 cases), colorectal (n = 835), gastric (n = 170), kidney (n = 143), lung (n = 332), prostate (n = 869), and urothelial (n = 428) cancer and B-cell lymphoma (n = 438). Case–control pairs were matched on age, sex, country of birth, and smoking (lung cancer study). Conditional logistic regression was used to assess associations between cancer risk and methylation at 45 CpGs of FOXO3 included on the HumanMethylation450 assay. Mixed-effects Cox models were used to estimate hazard ratios (HR) and 95% confidence intervals (CI) for associations with cancer survival (total n = 2286 deaths). Additionally, using data from 1088 older participants, we assessed associations of FOXO3 methylation with overall and cause-specific mortality (n = 354 deaths). Methylation at a CpG in the first exon region of FOXO3 (6:108882981) was associated with gastric cancer survival (HR = 2.39, 95% CI: 1.60–3.56, p = 1.9 × 10−5). Methylation at three CpGs in TSS1500 and gene body was associated with lung cancer survival (p < 6.1 × 10−5). We found no evidence of associations of FOXO3 methylation with cancer risk and mortality. Our findings may contribute to understanding the implication of FOXO3 in longevity.

Download Full-text

Logistic Regression for Matched Case-Control Studies

Applied Logistic Regression ◽

10.1002/0471722146.ch7 ◽

2005 ◽

pp. 223-259 ◽

Cited By ~ 2

Keyword(s):

Logistic Regression ◽

Case Control ◽

Case Control Studies ◽

Matched Case

Download Full-text

Penalized logistic regression with low prevalence exposures beyond high dimensional settings

PLoS ONE ◽

10.1371/journal.pone.0217057 ◽

2019 ◽

Vol 14 (5) ◽

pp. e0217057 ◽

Cited By ~ 10

Author(s):

Sam Doerken ◽

Marta Avalos ◽

Emmanuel Lagarde ◽

Martin Schumacher

Keyword(s):

Logistic Regression ◽

High Dimensional ◽

Penalized Logistic Regression ◽

Low Prevalence

Download Full-text

Sparse estimation for case–control studies with multiple disease subtypes

Biostatistics ◽

10.1093/biostatistics/kxz063 ◽

2020 ◽

Author(s):

Nadim Ballout ◽

Cedric Garcia ◽

Vivian Viallon

Keyword(s):

Logistic Regression ◽

Regression Model ◽

Logistic Regression Model ◽

Multinomial Logistic Regression ◽

Conditional Logistic Regression ◽

Case Control ◽

Case Control Studies ◽

Conditional Logistic Regression Model ◽

Symmetric Formulation ◽

Disease Subtypes

Summary The analysis of case–control studies with several disease subtypes is increasingly common, e.g. in cancer epidemiology. For matched designs, a natural strategy is based on a stratified conditional logistic regression model. Then, to account for the potential homogeneity among disease subtypes, we adapt the ideas of data shared lasso, which has been recently proposed for the estimation of stratified regression models. For unmatched designs, we compare two standard methods based on $L_1$-norm penalized multinomial logistic regression. We describe formal connections between these two approaches, from which practical guidance can be derived. We show that one of these approaches, which is based on a symmetric formulation of the multinomial logistic regression model, actually reduces to a data shared lasso version of the other. Consequently, the relative performance of the two approaches critically depends on the level of homogeneity that exists among disease subtypes: more precisely, when homogeneity is moderate to high, the non-symmetric formulation with controls as the reference is not recommended. Empirical results obtained from synthetic data are presented, which confirm the benefit of properly accounting for potential homogeneity under both matched and unmatched designs, in terms of estimation and prediction accuracy, variable selection and identification of heterogeneities. We also present preliminary results from the analysis of a case–control study nested within the EPIC (European Prospective Investigation into Cancer and nutrition) cohort, where the objective is to identify metabolites associated with the occurrence of subtypes of breast cancer.

Download Full-text

Prediction and Variable Selection in High-Dimensional Misspecified Binary Classification

Entropy ◽

10.3390/e22050543 ◽

2020 ◽

Vol 22 (5) ◽

pp. 543 ◽

Cited By ~ 2

Author(s):

Konrad Furmańczyk ◽

Wojciech Rejchel

Keyword(s):

Logistic Regression ◽

Variable Selection ◽

Logistic Model ◽

Binary Classification ◽

Model Misspecification ◽

High Dimensional ◽

Classification Models ◽

Computationally Efficient ◽

Class Labels ◽

Penalized Logistic Regression

In this paper, we consider prediction and variable selection in the misspecified binary classification models under the high-dimensional scenario. We focus on two approaches to classification, which are computationally efficient, but lead to model misspecification. The first one is to apply penalized logistic regression to the classification data, which possibly do not follow the logistic model. The second method is even more radical: we just treat class labels of objects as they were numbers and apply penalized linear regression. In this paper, we investigate thoroughly these two approaches and provide conditions, which guarantee that they are successful in prediction and variable selection. Our results hold even if the number of predictors is much larger than the sample size. The paper is completed by the experimental results.

Download Full-text