Learning complex dependency structure of gene regulatory networks from high dimensional micro-array data with Gaussian Bayesian networks

TL;DR

This paper compares a score-based Hill Climbing algorithm with Glasso for learning complex gene regulatory networks from high-dimensional micro-array data, demonstrating HC's superior accuracy and efficiency in capturing intricate dependencies.

Contribution

It introduces the use of a simple Hill Climbing algorithm for Gaussian Bayesian Network learning and compares its performance to Glasso in reconstructing gene regulatory networks.

Findings

HC learns dependencies more accurately and efficiently.

HC models complex local and global gene interactions.

Glasso tends to model unnecessary dependencies and introduces structural bias.

Abstract

Gene expression datasets consist of thousand of genes with relatively small samplesizes (i.e. are large-$p$-small-$n$). Moreover, dependencies of various orders co-exist in the datasets. In the Undirected probabilistic Graphical Model (UGM) framework the Glasso algorithm has been proposed to deal with high dimensional micro-array datasets forcing sparsity. Also, modifications of the default Glasso algorithm are developed to overcome the problem of complex interaction structure. In this work we advocate the use of a simple score-based Hill Climbing algorithm (HC) that learns Gaussian Bayesian Networks (BNs) leaning on Directed Acyclic Graphs (DAGs). We compare HC with Glasso and its modifications in the UGM framework on their capability to reconstruct GRNs from micro-array data belonging to the Escherichia Coli genome. We benefit from the analytical properties of the Joint Probability Density (JPD) function on which both directed and undirected PGMs build to convert DAGs to UGMs. We conclude that dependencies in complex data are learned best by the HC algorithm, presenting them most accurately and efficiently, simultaneously modelling strong local and weaker but significant global connections coexisting in the gene expression dataset. The HC algorithm adapts intrinsically to the complex dependency structure of the dataset, without forcing a specific structure in advance. On the contrary, Glasso and modifications model unnecessary dependencies at the expense of the probabilistic information in the network and of a structural bias in the JPD function that can only be relieved including many parameters.