Systematic identification of functional plant modules through the integration of complementary data sources

Area of Science:

• Plant genomics and functional genomics.
• Computational biology and bioinformatics.

Background:

• Understanding gene interactions and regulation is crucial for deciphering biological functions.
• Gene networks offer a powerful framework for identifying functional gene modules.

Purpose of the Study:

• To integrate diverse genome-wide functional genomics data to construct gene networks in Arabidopsis (Arabidopsis thaliana).
• To identify functional gene modules and predict functions for genes with unknown roles.
• To assess the conservation of these modules across different plant species.

Main Methods:

• Integration of large-scale expression data, functional gene annotations, protein-protein interactions, and transcription factor-target interactions.
• Construction of gene networks and delineation of gene modules.
• Comparative analysis of module conservation across plant species.
• Module-based functional prediction and experimental validation.

Main Results:

• 1,563 gene modules covering 13,142 genes were identified, with most showing strong coexpression.
• Highly connected hub genes were enriched for embryo lethality and biological process crosstalk.
• 58% of modules exhibited conserved coexpression across multiple plant species.
• Module-based predictions successfully inferred functions for 38.1% of 197 recently characterized genes.
• New functional hypotheses were generated for 1,701 Arabidopsis genes and 43,621 genes in six other plant species.

Conclusions:

• Combining multiple data types is advantageous for studying gene function and regulation.
• Module-based functional predictions provide a valuable tool for annotating genes with unknown functions.
• Inferred modules offer insights into coexpression and coregulation conservation, facilitating comparative functional genomics.