Inferring and analyzing gene regulatory networks from multi-factorial expression data: a complete and interactive...

Area of Science:

• Bioinformatics
• Computational Biology
• Genomics

Background:

• High-throughput transcriptomic datasets (RNA-seq) are crucial for identifying biological response actors and regulators.
• Existing graphical interfaces often lack advanced features like gene clustering or state-of-the-art regulatory network inference.
• There is a need for comprehensive tools that integrate advanced analytical methods for expression data.

Purpose of the Study:

• To develop an advanced, user-friendly interface for analyzing multi-factorial transcriptomic datasets.
• To provide state-of-the-art methods for gene clustering and regulatory network inference.
• To ensure reproducible and informative analysis of RNA-seq data.

Main Methods:

• Development of DIANE (Dashboard for the Inference and Analysis of Networks from Expression data) as an interactive workflow.
• Implementation of normalization, dimensionality reduction, differential expression, and ontology enrichment.
• Integration of Mixture Models for gene clustering and Random Forests for regulatory network inference with permutation-based significance assessment.

Main Results:

• DIANE offers a comprehensive suite of tools for analyzing complex expression datasets.
• The platform enables advanced gene clustering and robust gene regulatory network reconstruction.
• Novel statistical methods are included for assessing the significance of inferred regulatory influences.

Conclusions:

• DIANE effectively demonstrates the value of advanced analytical procedures for RNA-seq data exploration.
• The tool facilitates intuitive gene expression profile clustering and gene network reconstruction.
• DIANE provides a valuable resource for identifying candidate genes and signaling pathways, available as a web service or R package.