Recursive regularization for inferring gene networks from time-course gene expression profiles

Teppei Shimamura,Rui Yamaguchi,André Fujita,Satoru Miyano,Masao Nagasaki,Seiya Imoto

doi:10.1186/1752-0509-3-41

Teppei Shimamura, Rui Yamaguchi + Show 4 more

Open Access

https://doi.org/10.1186/1752-0509-3-41

Copy DOI

Abstract

BackgroundInferring gene networks from time-course microarray experiments with vector autoregressive (VAR) model is the process of identifying functional associations between genes through multivariate time series. This problem can be cast as a variable selection problem in Statistics. One of the promising methods for variable selection is the elastic net proposed by Zou and Hastie (2005). However, VAR modeling with the elastic net succeeds in increasing the number of true positives while it also results in increasing the number of false positives.ResultsBy incorporating relative importance of the VAR coefficients into the elastic net, we propose a new class of regularization, called recursive elastic net, to increase the capability of the elastic net and estimate gene networks based on the VAR model. The recursive elastic net can reduce the number of false positives gradually by updating the importance. Numerical simulations and comparisons demonstrate that the proposed method succeeds in reducing the number of false positives drastically while keeping the high number of true positives in the network inference and achieves two or more times higher true discovery rate (the proportion of true positives among the selected edges) than the competing methods even when the number of time points is small. We also compared our method with various reverse-engineering algorithms on experimental data of MCF-7 breast cancer cells stimulated with two ErbB ligands, EGF and HRG.ConclusionThe recursive elastic net is a powerful tool for inferring gene networks from time-course gene expression profiles.

Highlights

Inferring gene networks from time-course microarray experiments with vector autoregressive (VAR) model is the process of identifying functional associations between genes through multivariate time series
The recursive elastic net is a powerful tool for inferring gene networks from timecourse gene expression profiles
To increase the capability of the elastic net, i.e., to decrease the number of false positives while keeping the high number of true positives in inferring gene networks based on VAR model, we propose a new class of regularization, called recursive elastic net, by incorporating relative importance of the coefficients

Summary

Introduction

Inferring gene networks from time-course microarray experiments with vector autoregressive (VAR) model is the process of identifying functional associations between genes through multivariate time series. This problem can be cast as a variable selection problem in Statistics. The inference of gene networks from time-course microarray data can be defined as the process of identifying functional interactions between genes over time. We use vector autoregressive (VAR) model [6,7] to estimate gene networks from time-course micro-. The process of inferring gene networks based on the VAR model is to choose non-zero coefficients in the coefficient matrix, which can be considered as a problem of statistical model selection, especially as a variable selection problem [8]. A variety of variable selection methods have been developed, e.g., best-subset selection [9], subset selection [9] and the lasso [10], these methods often suffer from the following crucial problems due to the limited number of samples (time points) compared with the large number of variables (genes) in timecourse microarray data

Objectives

Methods

Results

Conclusion