Non-homogeneous dynamic Bayesian networks with Bayesian regularization for inferring gene regulatory networks with gradually time-varying structure

Authors:
Frank Dondelinger;Sophie Lèbre;Dirk Husmeier
Affiliations:
Biomathematics and Statistics Scotland, JCMB, Edinburgh, UK EH9 3JZ and Institute for Adaptive and Neural Computation, The University of Edinburgh, Edinburgh, UK EH8 9AB;LSIIT, UMR 7005, Université de Strasbourg, Illkirch, France 67412;Biomathematics and Statistics Scotland, JCMB, Edinburgh, UK EH9 3JZ and School of Mathematics and Statistics, University of Glasgow, Glasgow, UK G12 8QW
Venue:
Machine Learning
Year:
2013

Citing 9
Cited 1

The relationship between Precision-Recall and ROC curves

ICML '06 Proceedings of the 23rd international conference on Machine learning
Inferring gene regulatory networks from time series data using the minimum description length principle

Bioinformatics
Recovering temporally rewiring networks: a model-based approach

Proceedings of the 24th international conference on Machine learning
Modeling changing dependency structure in multivariate time series

Proceedings of the 24th international conference on Machine learning
Learning Non-Stationary Dynamic Bayesian Networks

The Journal of Machine Learning Research
Non-homogeneous dynamic Bayesian networks for continuous data

Machine Learning
Joint Bayesian model selection and estimation of noisy sinusoidsvia reversible jump MCMC

IEEE Transactions on Signal Processing
Bayesian curve fitting using MCMC with applications to signalsegmentation

IEEE Transactions on Signal Processing
Time Varying Dynamic Bayesian Network for Nonstationary Events Modeling and Online Inference

IEEE Transactions on Signal Processing

Regularization of non-homogeneous dynamic Bayesian networks with global information-coupling based on hierarchical Bayesian models

Machine Learning

Quantified Score

Hi-index	0.00

Visualization

Abstract

The proper functioning of any living cell relies on complex networks of gene regulation. These regulatory interactions are not static but respond to changes in the environment and evolve during the life cycle of an organism. A challenging objective in computational systems biology is to infer these time-varying gene regulatory networks from typically short time series of transcriptional profiles. While homogeneous models, like conventional dynamic Bayesian networks, lack the flexibility to succeed in this task, fully flexible models suffer from inflated inference uncertainty due to the limited amount of available data. In the present paper we explore a semi-flexible model based on a piecewise homogeneous dynamic Bayesian network regularized by gene-specific inter-segment information sharing. We explore different choices of prior distribution and information coupling and evaluate their performance on synthetic data. We apply our method to gene expression time series obtained during the life cycle of Drosophila melanogaster, and compare the predicted segmentation with other state-of-the-art techniques. We conclude our evaluation with an application to synthetic biology, where the objective is to predict an in vivo regulatory network of five genes in Saccharomyces cerevisiae subjected to a changing environment.