A dynamic programing approach to integrate gene expression data and network information for pathway model generation.

Area of Science:

• Bioinformatics
• Systems Biology
• Computational Biology

Background:

• Increasing volumes of biological data necessitate advanced bioinformatics methods for integration.
• Existing methods for identifying active biological modules often lack interpretability for pathway modeling.
• Translating identified modules into actionable insights for drug design and treatment remains a challenge.

Purpose of the Study:

• To develop a novel algorithm, IMPRes, for step-wise active pathway detection.
• To improve the interpretability of bioinformatics results for pathway generation and testing.
• To leverage existing pathway knowledge and omics data for enhanced biological discovery.

Main Methods:

• Developed the IMPRes algorithm employing a dynamic programming approach.
• Integrated pathway interaction knowledge from the Kyoto Encyclopedia of Genes and Genomes.
• Utilized omics data to assign penalties to genes, interactions, and pathways.
• Applied a shortest path algorithm to detect downstream pathways from seed genes.

Main Results:

• IMPRes demonstrated competitive or superior performance compared to state-of-the-art methods on three yeast datasets.
• A case study on human lung cancer data revealed novel insights into involved genes and mechanisms.
• The step-wise detection facilitates pathway tracing, potentially improving drug design and treatment strategies.

Conclusions:

• IMPRes offers a robust and interpretable method for active pathway detection from omics data.
• The algorithm provides valuable insights for understanding complex biological systems and diseases.
• IMPRes is accessible via a web server, promoting its use in biological research.