arXiv subjects

Karim Oualkacha

Publications and source records attributed to Karim Oualkacha.

13 recordsLinked to original sources

arXiv2026

Global testing of SNP-methylation interactions on binary phenotypes via a logistic functional regression model

Understanding how genetic and epigenetic factors jointly influence binary health outcomes remains a major challenge in biomedical research. We propose a global test for the overall effect of interactions between DNA methylation and a set of single nucleotide polymorphisms (SNPs) on a binary phenotype. We propose a logistic functional regression model in which methylation measurements at CpG sites are transformed into smooth functional predictors interacting with discrete SNP genotypes through a localized kernel. This framework enables stable inference on region-level interactions while accounting for the spatial structure of methylation around SNPs. Extensive simulations show that the proposed test provides well-calibrated type I error and improved power over classical SNP-CpG pairwise analyses. The practical relevance of the method is illustrated using publicly available methylation and genotyping data from an obesity case-control study.

stat.ME↗

arXiv2026

Incomplete Matrix Regression

Matrix completion seeks to recover a low-rank matrix from a sparse and noisy subset of its entries. In many applications, such as recommendation systems and urban mobility, the observed matrix is accompanied by auxiliary covariates on its rows and columns and exhibits dependence across them. We propose Incomplete Matrix Regression (IMR), a distribution-free penalized regression framework that integrates such information into matrix completion. The target matrix is modeled as the sum of intercepts, covariate effects regularized by a Lasso penalty, and a low-rank latent component that captures structure unexplained by the covariates. Known similarity structures, such as spatial and temporal kernels, are incorporated through ridge-type penalties on the latent factors. For estimation, we provide a scalable alternating least-squares algorithm whose modular form allows us to include or exclude individual model components without rederiving the updates. We establish non-asymptotic error bounds for both the Lasso and matrix completion estimators that are consistent with standard rates in their respective literature. Through simulation studies and two real-data applications, we demonstrate that the proposed method attains predictive accuracy competitive with more complex methods at a small fraction of their computational cost. The methodology is implemented in the R package IMR.

stat.ME↗

arXiv2025

A Functional Approach to Testing Overall Effect of Interaction Between DNA Methylation and SNPs

We introduce a test for the overall effect of interaction between DNA methylation and a set of single nucleotide polymorphisms (SNPs) on a quantitative phenotype. The developed inference procedure is based on a functional approach that extends existing regression models in functional data analysis. Through extensive simulations, we show that the proposed test effectively controls type I error rates and highlights increased empirical power over existing methods, particularly when multiple interactions are present. The use of the proposed test is illustrated with an application to data from obesity patients and controls.

stat.ME↗

arXiv2025

An efficient hybrid approach of quantile and expectile regression

Quantiles and expectiles are determined by different loss functions: asymmetric least absolute deviation for quantiles and asymmetric squared loss for expectiles. This distinction ensures that quantile regression methods are robust to outliers but somewhat less effective than expectile regression, especially for normally distributed data. However, expectile regression is vulnerable to lack of robustness, especially for heavy-tailed distributions. To address this trade-off between robustness and effectiveness, we propose a novel approach. By introducing a parameter $γ$ that ranges between 0 and 1, we combine the aforementioned loss functions, resulting in a hybrid approach of quantiles and expectiles. This fusion leads to the estimation of a new type of location parameter family within the linear regression framework, termed Hybrid of Quantile and Expectile Regression (HQER). The asymptotic properties of the resulting estimaror are then established. Through simulation studies, we compare the asymptotic relative efficiency of the HQER estimator with its competitors, namely the quantile, expectile, and $k$th power expectile regression estimators. Our results show that HQER outperforms its competitors in several simulation scenarios. In addition, we apply HQER to a real dataset to illustrate its practical utility.

stat.ME↗

arXiv2025

Penalized generalized linear mixed models for longitudinal outcomes in genetic association studies

This work is motivated by analyses of longitudinal data collected from participants in the Quebec Longitudinal Study of Child Development (QLSCD) and the Quebec Newborn Twin Study (QNTS) to identify important genetic predictors for emotional and behavioral difficulties in childhood and adolescence. We propose a lasso penalized mixed model for continuous and binary longitudinal traits that allows the inclusion of multiple random effects to account for random individual effects not attributable to the genetic similarity between individuals. Through simulation studies, we show that replacing the estimated genetic relatedness matrix (GRM) by a sparse matrix introduces bias in the variance components estimates, but that the obtained computational gain is major while the impact on the performance of the penalized model to retrieve important predictors is negligible. We compare the performance of the proposed penalized mixed model to a standard lasso and to a univariate mixed model association test and show that the proposed model always identifies causal predictors with greater precision. Finally, we show an application of the proposed methodology to predict three externalizing behavorial scores in the combined QLSCD and QNTS longitudinal cohorts.

stat.ME↗

arXiv2023

Hierarchical selection of genetic and gene by environment interaction effects in high-dimensional mixed models

Interactions between genes and environmental factors may play a key role in the etiology of many common disorders. Several regularized generalized linear models (GLMs) have been proposed for hierarchical selection of gene by environment interaction (GEI) effects, where a GEI effect is selected only if the corresponding genetic main effect is also selected in the model. However, none of these methods allow to include random effects to account for population structure, subject relatedness and shared environmental exposure. In this paper, we develop a unified approach based on regularized penalized quasi-likelihood (PQL) estimation to perform hierarchical selection of GEI effects in sparse regularized mixed models. We compare the selection and prediction accuracy of our proposed model with existing methods through simulations under the presence of population structure and shared environmental exposure. We show that for all simulation scenarios, compared to other penalized methods, our proposed method enforced sparsity by controlling the number of false positives in the model while having the best predictive performance. Finally, we apply our method to a real data application using the Orofacial Pain: Prospective Evaluation and Risk Assessment (OPPERA) study, and found that our method retrieves previously reported significant loci.

stat.ME↗

arXiv2023

High-Dimensional Penalized Bernstein Support Vector Machines

The support vector machines (SVM) is a powerful classifier used for binary classification to improve the prediction accuracy. However, the non-differentiability of the SVM hinge loss function can lead to computational difficulties in high dimensional settings. To overcome this problem, we rely on Bernstein polynomial and propose a new smoothed version of the SVM hinge loss called the Bernstein support vector machine (BernSVM), which is suitable for the high dimension $p >> n$ regime. As the BernSVM objective loss function is of the class $C^2$, we propose two efficient algorithms for computing the solution of the penalized BernSVM. The first algorithm is based on coordinate descent with maximization-majorization (MM) principle and the second one is IRLS-type algorithm (iterative re-weighted least squares). Under standard assumptions, we derive a cone condition and a restricted strong convexity to establish an upper bound for the weighted Lasso BernSVM estimator. Using a local linear approximation, we extend the latter result to penalized BernSVM with non convex penalties SCAD and MCP. Our bound holds with high probability and achieves a rate of order $\sqrt{s\log(p)/n}$, where $s$ is the number of active features. Simulation studies are considered to illustrate the prediction accuracy of BernSVM to its competitors and also to compare the performance of the two algorithms in terms of computational timing and error estimation. The use of the proposed method is illustrated through analysis of three large-scale real data examples.

stat.ML↗

arXiv2022

Efficient Penalized Generalized Linear Mixed Models for Variable Selection and Genetic Risk Prediction in High-Dimensional Data

Sparse regularized regression methods are now widely used in genome-wide association studies (GWAS) to address the multiple testing burden that limits discovery of potentially important predictors. Linear mixed models (LMMs) have become an attractive alternative to principal components (PC) adjustment to account for population structure and relatedness in high-dimensional penalized models. However, their use in binary trait GWAS rely on the invalid assumption that the residual variance does not depend on the estimated regression coefficients. Moreover, LMMs use a single spectral decomposition of the covariance matrix of the responses, which is no longer possible in generalized linear mixed models (GLMMs). We introduce a new method called pglmm, a penalized GLMM that allows to simultaneously select genetic markers and estimate their effects, accounting for between-individual correlations and binary nature of the trait. We develop a computationally efficient algorithm based on PQL estimation that allows to scale regularized mixed models on high-dimensional binary trait GWAS (~300,000 SNPs). We show through simulations that penalized LMM and logistic regression with PC adjustment fail to correctly select important predictors and/or that prediction accuracy decreases for a binary response when the dimensionality of the relatedness matrix is high compared to pglmm. Further, we demonstrate through the analysis of two polygenic binary traits in the UK Biobank data that our method can achieve higher predictive performance, while also selecting fewer predictors than a sparse regularized logistic lasso with PC adjustment. Our method is available as a Julia package PenalizedGLMM.jl.

stat.ME↗

arXiv2021

A copula-based set-variant association test for bivariate continuous or mixed phenotypes

In genome wide association studies (GWAS), researchers are often dealing with non-normally distributed traits or a mixture of discrete-continuous traits. However, most of the current region-based methods rely on multivariate linear mixed models (mvLMMs) and assume a multivariate normal distribution for the phenotypes of interest. Hence, these methods are not applicable to disease or non-normally distributed traits. Therefore, there is a need to develop unified and flexible methods to study association between a set of (possibly rare) genetic variants and non-normal multivariate phenotypes. Copulas are multivariate distribution functions with uniform margins on the $[0, 1]$ interval and they provide suitable models to deal with non-normality of errors in multivariate association studies. We propose a novel unified and flexible Copula-Based Multivariate Association Test (CBMAT) for discovering association between a genetic region and a bivariate continuous or mixed phenotype. We also derive a data-driven analytic p-value procedure of the proposed region-based score-type test. Through simulation studies, we demonstrate that CBMAT has well controlled type I error rates and higher power to detect associations compared with other existing methods, for discrete and non-normally distributed traits. At last, we apply CBMAT to detect the association between two genes located on chromosome 11 and several lipid levels measured on 1,477 subjects from the ASLPAC study.

stat.ME↗

arXiv2021

Weighted asymmetric least squares regression with fixed-effects

The fixed-effects model estimates the regressor effects on the mean of the response, which is inadequate to summarize the variable relationships in the presence of heteroscedasticity. In this paper, we adapt the asymmetric least squares (expectile) regression to the fixed-effects model and propose a new model: expectile regression with fixed-effects $(\ERFE).$ The $\ERFE$ model applies the within transformation strategy to concentrate out the incidental parameter and estimates the regressor effects on the expectiles of the response distribution. The $\ERFE$ model captures the data heteroscedasticity and eliminates any bias resulting from the correlation between the regressors and the omitted factors. We derive the asymptotic properties of the $\ERFE$ estimators and suggest robust estimators of its covariance matrix. Our simulations show that the $\ERFE$ estimator is unbiased and outperforms its competitors. Our real data analysis shows its ability to capture data heteroscedasticity (see our R package, \url{github.com/AmBarry/erfe}).

econ.EM↗

arXiv2021

Detecting differentially methylated regions in bisulfite sequencing data using quasi-binomial mixed models with smooth covariate effect estimates

Identifying disease-associated changes in DNA methylation can help to gain a better understanding of disease etiology. Bisulfite sequencing technology allows the generation of methylation profiles at single base of DNA. We previously developed a method for estimating smooth covariate effects and identifying differentially methylated regions (DMRs) from bisulfite sequencing data, which copes with experimental errors and variable read depths; this method utilizes the binomial distribution to characterize the variability in the methylated counts. However, bisulfite sequencing data frequently include low-count integers and can exhibit over or under dispersion relative to the binomial distribution. We present a substantial improvement to our previous work by proposing a quasi-likelihood-based regional testing approach which accounts for multiplicative and additive sources of dispersion. We demonstrate the theoretical properties of the resulting tests, as well as their marginal and conditional interpretations. Simulations show that the proposed method provides correct inference for smooth covariate effects and captures the major methylation patterns with excellent power.

stat.ME↗

arXiv2020

A new GEE method to account for heteroscedasticity, using asymmetric least-square regressions

Generalized estimating equations (GEE) are widely used to analyze longitudinal data; however, they are not appropriate for heteroscedastic data, because they only estimate regressor effects on the mean response{\textemdash}and therefore do not account for data heterogeneity. Here, we combine the GEE with the asymmetric least squares (expectile) regression to derive a new class of estimators, which we call generalized expectile estimating equations (GEEE). The GEEE model estimates regressor effects on the expectiles of the response distribution, which provides a detailed view of regressor effects on the entire response distribution. In addition to capturing data heteroscedasticity, the GEEE extends the various working correlation structures to account for within-subject dependence. We derive the asymptotic properties of the GEEE estimators and propose a robust estimator of its covariance matrix for inference (see our R package, github.com/AmBarry/expectgee). Our simulations show that the GEEE estimator is non-biased and efficient, and our real data analysis shows it captures heteroscedasticity.

stat.ME↗

arXiv2020

The Kendall Interaction Filter for Variable Interaction Screening in Ultra High Dimensional Classification Problems

Accounting for important interaction effects can improve prediction of many statistical learning models. Identification of relevant interactions, however, is a challenging issue owing to their ultrahigh-dimensional nature. Interaction screening strategies can alleviate such issues. However, due to heavier tail distribution and complex dependence structure of interaction effects, innovative robust and/or model-free methods for screening interactions are required to better scale analysis of complex and high-throughput data. In this work, we develop a new model-free interaction screening method, termed Kendall Interaction Filter (KIF), for the classification in high-dimensional settings. The KIF method suggests a weighted-sum measure, which compares the overall to the within-cluster Kendall's $τ$ of pairs of predictors, to select interactive couples of features. The proposed KIF measure captures relevant interactions for the clusters response-variable, handles continuous, categorical or a mixture of continuous-categorical features, and is invariant under monotonic transformations. We show that the KIF measure enjoys the sure screening property in the high-dimensional setting under mild conditions, without imposing sub-exponential moment assumptions on the features' distributions. We illustrate the favorable behavior of the proposed methodology compared to the methods in the same category using simulation studies, and we conduct real data analyses to demonstrate its utility.

stat.ME↗