Study of gene function based on spatial co-expression in a high-resolution mouse brain atlas

Area of Science:

• Neuroscience
• Bioinformatics
• Computational Biology

Background:

• The Allen Brain Atlas (ABA) provides high-resolution 3D gene expression data for thousands of genes in mouse brains.
• Analyzing this vast dataset is challenging due to its complexity and volume.
• Current ABA database use is limited to visual inspection, with unexplored potential for computational data mining.

Purpose of the Study:

• To test the hypothesis that genes with similar 3D expression profiles in the mouse brain share similar biological functions.
• To develop computational tools for efficient data mining of the ABA database.

Main Methods:

• Developed the histogram-row-column (HRC) algorithm for robust image filtering and pattern comparison.
• Implemented a semi-automatic approach for identifying genes with similar in situ hybridization patterns.
• Utilized the HRC algorithm for automatic pattern searching within the ABA image collection.

Main Results:

• The HRC algorithm successfully identifies manageable sets of gene pairs with similar expression patterns.
• Discovered gene expression patterns in proximity to genes of similar functional categories.
• Demonstrated the HRC algorithm's sensitivity in detecting subtle expression similarities.

Conclusions:

• The HRC algorithm is a fully automated tool for rapidly mining large brain image databases like ABA.
• Identified genes with similar spatial co-distribution patterns can be used for further functional investigation.
• 3D in situ hybridization patterns can serve as functional fingerprints for genes, leveraging ABA data with tools like HRC.