The functional landscape of mouse gene expression.

Area of Science:

• Genomics
• Molecular Biology
• Bioinformatics

Background:

• Transcriptional co-expression analysis is established for dissecting regulatory networks in model organisms.
• Tissue-specific gene expression is thought to indicate function in mammals but lacks objective testing.
• This study compares co-expression with tissue specificity for predicting mammalian gene function.

Purpose of the Study:

• To objectively test and compare the predictive power of gene co-expression versus tissue-specific expression for determining gene function in mammals.
• To establish quantitative transcriptional co-expression as a robust strategy for mammalian functional genomics.

Main Methods:

• Generated microarray expression data for ~40,000 mRNAs across 55 mouse tissues using custom oligonucleotide arrays.
• Analyzed gene expression patterns to identify correlations between co-expression and functional categories (Gene Ontology Biological Processes).

Main Results:

• Quantitative transcriptional co-expression strongly predicts gene function across diverse biological processes.
• Hundreds of functional categories show characteristic expression patterns, irrespective of tissue of origin.
• Tissue-specific expression is a weak predictor of gene function compared to co-expression.
• The gene PWP1, widely expressed, co-expresses with RNA-processing genes, indicating its role in rRNA biogenesis.

Conclusions:

• Functional genomics strategies using quantitative transcriptional co-expression are highly effective in mammals.
• Mammalian physiological control exhibits a more modular transcriptional regulation than previously understood.
• The study provides a public resource for mammalian functional genomics research.