Linking tissues to phenotypes using gene expression profiles.

Area of Science:

• Genetics and Genomics
• Systems Biology
• Bioinformatics

Background:

• Many human heritable diseases lack identified genetic causes, hindering effective treatment and research.
• Model organisms provide valuable data for disease gene discovery, but data are fragmented across databases.
• Understanding gene-phenotype relationships requires integrating diverse biological information.

Purpose of the Study:

• To create an integrated data resource linking mouse tissue gene expression with phenotypic data.
• To explore the relationship between gene expression location and phenotype manifestation across tissues.
• To enhance the identification of disease-causing genes by incorporating tissue-specific information.

Main Methods:

• Developed an integrated database combining mouse gene expression and phenotype data.
• Performed automated evaluation of predicted tissue-phenotype associations.
• Assessed the utility of the resource in improving disease-gene association prediction.

Main Results:

• 72-76% of phenotypes were associated with genes expressed in the affected tissue.
• 55-64% of phenotype-tissue associations showed spatially separated gene expression and phenotype manifestation.
• Integrating tissue-phenotype associations improved the detection of known disease-gene links, exemplified by JAK2 for Familial Erythrocytosis 1.

Conclusions:

• The integrated resource advances understanding of molecular-to-phenotype consequence chains.
• Tissue expression data is crucial for uncovering complex gene-phenotype relationships, even when spatially separated.
• This approach aids in identifying genetic causes for human diseases and understanding disease mechanisms.