Inference of regulatory networks from microarray data with R and the bioconductor package qpgraph

Area of Science:

• Bioinformatics
• Systems Biology
• Computational Biology

Background:

• Microarray data analysis aims to infer functional gene interactions.
• Distinguishing direct from indirect regulatory interactions is a key challenge.
• High dimensionality of microarray data (p >> n) limits standard multivariate methods.

Purpose of the Study:

• To present an intuitive method for inferring molecular regulatory networks from microarray data.
• To address the challenge of distinguishing direct from indirect gene interactions.
• To illustrate the application of this method using the R package qpgraph.

Main Methods:

• Utilizing limited-order partial correlations to infer network structures.
• Applying multivariate statistical approaches adapted for high-dimensional data.
• Demonstrating the method with the qpgraph package within the Bioconductor project.

Main Results:

• The proposed method offers an intuitive approach to network inference.
• It effectively tackles the challenge of distinguishing direct vs. indirect interactions.
• The qpgraph R package provides a practical implementation for biologists.

Conclusions:

• Limited-order partial correlations offer a viable solution for inferring gene regulatory networks from high-dimensional microarray data.
• The qpgraph package facilitates the application of this method in biological research.
• This approach enhances the accuracy of functional interaction blueprints derived from gene expression data.