Joint Modeling of An Outcome Variable and Integrated Omic Datasets Using GLM-PO2PLS

TL;DR

This paper introduces a novel joint modeling approach for an outcome variable and multiple omic datasets, enhancing understanding of their combined effects in disease studies through a new dimension reduction method.

Contribution

It extends existing dimension reduction techniques to jointly model outcomes and omics, providing a new method with proven identifiability and inference procedures.

Findings

The method effectively captures joint omic-outcome relationships.

Simulation studies validate the model's performance.

Application to Down syndrome data demonstrates practical utility.

Abstract

In many studies of human diseases, multiple omic datasets are measured. Typically, these omic datasets are studied one by one with the disease, thus the relationship between omics are overlooked. Modeling the joint part of multiple omics and its association to the outcome disease will provide insights into the complex molecular base of the disease. In this article, we extend dimension reduction methods which model the joint part of omics to a novel method that jointly models an outcome variable with omics. We establish the model identifiability and develop EM algorithms to obtain maximum likelihood estimators of the parameters for normally and Bernoulli distributed outcomes. Test statistics are proposed to infer the association between the outcome and omics, and their asymptotic distributions are derived. Extensive simulation studies are conducted to evaluate the proposed model. The model is illustrated by a Down syndrome study where Down syndrome and two omics - methylation and glycomics - are jointly modeled.