A deep neural network based regression model for triglyceride concentrations prediction using epigenome-wide DNA methylation profiles

Md Mohaiminul Islam,Pingzhao Hu,Ye Tian,Yang Wang,Yan Cheng

doi:10.1186/s12919-018-0121-1

Abstract

BackgroundEpigenetic modification has an effect on gene expression under the environmental alteration, but it does not change corresponding genome sequence. DNA methylation (DNAm) is one of the important epigenetic mechanisms. DNAm variations could be used as epigenetic markers to predict and account for the change of many human phenotypic traits, such as cancer, diabetes, and high blood pressure. In this study, we built deep neural network (DNN) regression models to account for interindividual variation in triglyceride concentrations measured at different visits of peripheral blood samples using epigenome-wide DNAm profiles.ResultsWe used epigenome-wide DNAm profiles of before and after medication interventions (called pretreatment and posttreatment, respectively) to predict triglyceride concentrations for peripheral blood draws at visit 2 (using pretreatment data) and at visit 4 (using both pretreatment and posttreatment data). Our experimental results showed that DNN models can predict triglyceride concentrations for blood draws at visit 4 using pretreatment and posttreatment DNAm data more accurately than for blood draws at visit 2 using pretreatment DNAm data. Furthermore, we got the best prediction results when we used pretreatment DNAm data to predict triglyceride concentrations for blood draws at visit 4, which suggests a long-term epigenetic effect on phenotypic traits. We compared the prediction performances of our proposed DNN models with that of support vector machine (SVM). This comparison showed that our DNN models achieved better prediction performance than did SVM.ConclusionsWe demonstrated the superiority of our proposed DNN models over the SVM model for predicting triglyceride concentrations. This study also suggests that the DNN approach has advantages over other traditional machine-learning methods to model high-dimensional epigenome-wide DNAm data and other genomic data.

Highlights

Epigenetic modification has an effect on gene expression under the environmental alteration, but it does not change corresponding genome sequence
The p values of the Shapiro test of the log of observed triglyceride concentrations in test sets from Case A, Case B, and Case C, were 0.17, 0.25, and 0.25, respectively, suggesting that the observed triglyceride levels followed log-normal distribution
We performed the same procedure on their averaged predicted values from the three splits of training and test sets using the support vector machine (SVM) models with largest Cor values and the deep neural network (DNN) model with largest Cor values and the p values for Case A, Case B, and Case C were 0.09, 0.05, and 0.78, respectively, for SVM models, and 0.08, 0.14, and 0.59, respectively, for DNN models, which suggest that the predicted triglyceride levels using either DNN models or SVM models follow log-normal distribution

Summary

Introduction

Epigenetic modification has an effect on gene expression under the environmental alteration, but it does not change corresponding genome sequence. DNAm variations could be used as epigenetic markers to predict and account for the change of many human phenotypic traits, such as cancer, diabetes, and high blood pressure. We built deep neural network (DNN) regression models to account for interindividual variation in triglyceride concentrations measured at different visits of peripheral blood samples using epigenome-wide DNAm profiles. Having a high concentration of triglycerides in human blood can increase our risk of heart, stroke, and other diseases. The development of new high-throughput genomic technologies makes it natural to extend these phenotypic prediction models to complex traits, such as triglyceride. Using DNAm profiles to predict disease phenotypic courses has not yet been explored in detail

Methods

Results

Conclusion