CORTx: A Functional Browser for the Transcriptomes of Neocortical Layers

This interface provides numerous additional ways to visualize and explore the dataset initially described in "A Transcriptomic Atlas of Mouse Neocortical Layers" (supplementary website).

Find correlated genes: Enter a single mouse gene symbol or gene ID used by Ensembl to sort other genes by expression correlation across all samples from all three sets of dissections.

Browse functional annotations by average significance of enrichment in a layer (must be enriched in all available dissections and have at least 3 associated genes in one dissection):
Layers 2/3, Layer 4, Layer 5, Layer 6, Layer 6b, and no layer enrichment.

Browse functional annotations

Search functional annotations: Enter keywords to search functional annotations. Search is case insensitive but word order matters.

Find functional annotations associated with a gene: Enter a single mouse gene symbol or gene ID used by Ensembl.

Layer views of the functional annotations highlighted in the Neuron publication: Type I Diabetes, NMDA Receptor, Alzheimer's Disease, Parkinson's Disease, and Coronary Artery Disease.

Correspondence of the dissected samples to layers (red reflects higher expression):

The heatmaps were made with Matrix2png. Values reflect the calculation described in Figure 2 of the Neuron paper, though using only genes that were annotated to be specific markers of each layer.

Notes on interpretation: In the 'gene expression view', no FPKM is represented as being less than 0.05 even if it was quantified as less than that (even zero) since the y-axis is on a logarithmic scale. When interpreting the layer distributions, please bear in mind that layers 2/3 and 5 have far more genes than other layers. You can visualize these differences for many functional databases by comparing with the 'background' annotation (see below). Functions are sorted in each layer based on significance of enrichment of genes with that annotation over what would be expected by chance (average one-sided p-value across all three regions, as calculated by Fisher's exact test on a 2x2 contingency table, requiring at least three quantified genes to have that annotation in at least one region).

Sources of function annotations: autism candidate genes from SFARI GENE and Basu, et al.; coexpressed modules in mouse brain from Miller, et al.; coexpressed modules in mouse neurons from Winden, et al.; genes enriched in particular cell types from Cahoy, et al.; GO biological process, cellular component, and molecular function from Ensembl; KEGG pathways from, well, KEGG PATHWAY; mouse knockout phenotypes from MGI; mouse orthologs of human genes in GWAS loci translated from Ensembl Variation; Reactome complex/interaction/pathway from Reactome; and one-to-one mouse orthologs of schizophrenia candidate genes from NCGR. Translations of human genes to mouse and to Ensembl gene IDs were done using Ensembl-defined one-to-one orthologs (release 56) and the MGI batch query. 'Background' refers to the union of all genes associated with an annotation in the database, and can thus be used for comparison with layer distributions of genes having particular functions.

Please direct questions about the website to Grant Belgard. This site is not intended for programatic access.