Computational neuroanatomy and gene expression: Optimal sets of marker genes for brain regions

摘要

The three-dimensional data-driven Allen Gene Expression Atlas of the adultmouse brain consists of numerized in-situ hybridization data for thousands ofgenes, co-registered to the Allen Reference Atlas. We propose quantitativecriteria to rank genes as markers of a brain region, based on the localizationof the gene expression and on its functional fitting to the shape of theregion. These criteria lead to natural generalizations to sets of genes. Wefind sets of genes weighted with coefficients of both signs with almost perfectlocalization in all major regions of the left hemisphere of the brain, exceptthe pallidum. Generalization of the fitting criterion with positivityconstraint provides a lesser improvement of the markers, but requires sparsersets of genes.