Integrated Analysis of Anatomical Gene Expression Maps and Co-Expression Networks Using a Database, ViBrism

Detection of gene expression-anatomy association in biological structure is crucial for understanding its function based on the molecular and genetic/genomic information. Particularly in the mammalian brain where there are estimated 25,000 genes expressed, systematic and comprehensive quantification of the expression densities in the whole three-dimensional (3D) anatomical context is critical. The combinatorial number of randomly selected genes is more than the cell number in the brain, which assumes that non-random combinatorial gene expression underlies the formation of a wide variety of functional brain regions composed of multiple cells.
  To determine the association systematically, we have introduced a novel framework, Transcriptome Tomography, for spatially integrating comprehensive endogenous gene expression within an isotropic anatomical context. Using this rapid and unbiased 3D mapping technique, in the first instance, we have generated a dataset of 36,000 maps covering the whole mouse brain (ViBrism: http://vibrism.riken.jp/3dviewer/ex/index.html) and validated them against existing data with respect to the expression location and density (paper submitted).
  Here, we used an informatics approach to identify the combinatorial gene expression as a broad co-expression network. The gene network links covering the whole brain followed an inverse-power law and were rich in functional interaction and gene ontology terms. Developmentally conserved co-expression modules underlie the network structure. To demonstrate the relevance of the finding, we mined Huntington’s disease gene (Htt) and found a novel disease-related co-expression network containing genes potentially co-functioning with Htt in neural differentiation and modulating the disease specific differential vulnerability in brain regions.
  The maps are spatially isotropic and well suited to analysis in the standard space for brain-atlas databases, e.g. Waxholm Space (PLoS Comput Biol 2011, 7[2]: e1001065) as shown in the related poster by J. Boline et., al. Our time and cost effective framework　will facilitate research creating and using open-resources for a molecular-based understanding of complex structures.

A part of this work was conducted within the Waxholm Space Task Force of the International Neuroinformatics Coordinating Facility (INCF) Program on Digital Brain Atlasing. We thank the program members, particularly, R. Baldock, I. Zaslavsky, L.Ibanez and J. Boline.