Inferring Gene Regulatory Networks from RNA-seq Data Using Kernel Classification.

Area of Science:

• Computational Biology
• Systems Biology
• Genomics

Background:

• Gene regulatory networks (GRNs) are crucial for understanding cellular processes.
• Gene expression profiling is a key method for inferring gene regulators and targets within GRNs.
• Analyzing diverse datasets like RNA-seq and microarray data is essential for comprehensive GRN construction.

Purpose of the Study:

• To construct a gene regulatory network for the Saccharomyces cerevisiae genome.
• To integrate and analyze RNA-seq and microarray data across various experimental conditions.
• To develop and evaluate a computational pipeline for GRN analysis.

Main Methods:

• Development of a data analysis and preparation pipeline.
• Application of kernel classification models (one-class, two-class, rare event) for gene categorization.
• Assessment of normalization technique impacts on RNA-seq data performance.

Main Results:

• The developed pipeline demonstrates robust performance in analyzing gene expression data.
• Classification models effectively categorize genes, contributing to network inference.
• High statistical performance metrics achieved, including 99% recall and 98% AUC score.

Conclusions:

• The study highlights the effectiveness of classification methods in enhancing GRN comprehension.
• The findings offer novel insights into gene interactions within the yeast regulatory network.
• The developed pipeline provides a valuable tool for future GRN research.