Bridging the Gap between Connectome and Transcriptome

Recent technical advances have enabled the construction of brain-wide atlases measuring expression levels of thousands of genes across many different brain regions in several distinct species. These atlases offer new opportunities for identifying transcriptional correlates of spatially varying properties of the connectome. Broad spatial gradients dominate the transcriptional landscape of the brain, track variations in regional cellular architecture, microcircuitry, and inter-regional connectivity, and may represent a molecular signature of regional specialisation. More specific signatures of inter-regional connectivity and connectome topology are superimposed on these broad gradients and are conserved across different species. The recent construction of brain-wide gene expression atlases, which measure the transcriptional activity of thousands of genes in multiple anatomical locations, has made it possible to connect spatial variations in gene expression to distributed properties of connectome structure and function. These analyses have revealed that spatial patterning of gene expression and neuronal connectivity are closely linked, following broad spatial gradients that track regional variations in microcircuitry, inter-regional connectivity, and functional specialisation. Superimposed on these gradients are more specific associations between gene expression and connectome topology that appear conserved across diverse species and different resolution scales. These findings demonstrate the utility of brain-wide gene expression atlases for bridging the gap between molecular function and large-scale connectome organisation in health and disease. The recent construction of brain-wide gene expression atlases, which measure the transcriptional activity of thousands of genes in multiple anatomical locations, has made it possible to connect spatial variations in gene expression to distributed properties of connectome structure and function. These analyses have revealed that spatial patterning of gene expression and neuronal connectivity are closely linked, following broad spatial gradients that track regional variations in microcircuitry, inter-regional connectivity, and functional specialisation. Superimposed on these gradients are more specific associations between gene expression and connectome topology that appear conserved across diverse species and different resolution scales. These findings demonstrate the utility of brain-wide gene expression atlases for bridging the gap between molecular function and large-scale connectome organisation in health and disease. the area under a receiver operating characteristic (ROC) curve is a summary measure of the accuracy of a classification and can be interpreted as the average sensitivity over all possible values of specificity. specific DNA variant chosen for further analysis (e.g., to test for an association with phenotypic variability). similarity (often quantified as a Pearson or Spearman correlation) between gene expression vectors for a pair of tissue samples or brain regions. network of brain regions including posterior cingulate, medial prefrontal and lateral parietal cortices. degree to which a gene exhibits a consistent regional expression profile across the six donor brains comprising the AHBA. causal (directed) influence that one neuronal system exerts over another. first principal component of expression variance for a subset of genes, where the subset is typically defined as a module of a gene coexpression network. genetic loci that impact the expression levels of a gene. statistical dependence between neurophysiological signals acquired from distinct neural elements. hierarchical organisation of terms (most commonly biological functions, but also molecular functions and cellular components), to which genes are constantly being annotated based on the latest scientific evidence. analytical approach that tests for associations between phenotypic variability and markers of allelic variation distributed throughout the genome. a network module is a subset of densely interconnected nodes. Hierarchical modularity occurs when modules contain nested submodules over several resolution scales. method for quantifying gene expression that allows subcellular localization of a specific segment of nucleic acid within an entire histological section. number of divisions separating two cells from a common progenitor. method for simultaneously quantifying the expression levels of thousands of genes through hybridization of Cy3-labeled RNA in a tissue sample to a probe on a microarray chip that maps to a unique DNA sequence. inter-regional statistical dependence in some measure of brain morphopmetry, such as regional volume, cortical thickness, or microstructure. pattern of gene expression that persists through development into adulthood. management of network resources (connectivity, activity, etc.) to support robust, adaptive and cost-efficient performance. number of connections attached to a network node. multivariate method that explains a set of observed variables as a linear combination of a set of predictor variables. distribution of the connections of a node across different topological modules of a network. Nodes with high participation have an equal distribution of edges across modules. set of highly connected network nodes that are more strongly interconnected with each other than expected by chance. network topology characterised by more clustered connectivity and comparable average shortest path length (the average minimum number of edges between nodes) relative to a random graph. anatomical connections between pairs of neural elements. ratio of T1-weighted to T2-weighted magnetic resonance image contrast, obtained by dividing the former image by the latter, is thought to provide a noninvasive marker of regional grey matter myelin content. the topology of a network is the specific arrangement of pairwise connections between nodes, and is invariant to the physical embedding of the system. complete set of RNA molecules in a cell. Some transcriptional assays may only sample a subset of these molecules. metabolic cost of neuronal connectivity. Often approximated by the volume or distance of anatomical connections.