Single Cell Transcriptomic Profiling
|Table of Contents|
One of the goals of the Allen Institute’s Cell Types Program is to classify cortical neurons based on gene expression captured using single-cell RNA sequencing. To date, RNA sequencing has been performed on cells from the lateral geniculate nucleus (LGd or LGN), primary visual cortex (VISp or V1) and the anterior lateral motor area (ALM) of the young adult laboratory mouse. Enrichment of cells representing discrete neuronal subpopulations was achieved using Cre-driven labeling and cell selection by Fluorescence Activated Cell Sorting (FACS). More information on the methods used in this study can be found in the whitepapers located in Documentation.
The transcriptomics profiling data is available as downloadable files from the RNA-Seq tab on the Cell Types landing page.
Data collected from the LGd has a preliminary data visualization tool utilizing a heatmap to view possible cell classifications. Clicking on "RNASeq heatmap for LGd" will take you to that visualization.
Utilizing the Heatmap Viewer for LGd Cells
Searching is available using three methods: (1) #Gene Search, when looking for a specific gene of interest, (2) #Differential Search, to find enhanced gene expression when comparing cell type features and 3) #Correlative Search, to find cell features that exhibit similar gene expression to a "seed gene" selected from the results of a Gene or a Differential Search, or to find genes that are co-expressed in specific cells.
You can take advantage of our or curated searches by clicking on one of the links from the landing page. For the broad class differential searches, the cell features chosen for the contrast cell features are all other broad classes. For the sub-class differential searches, the cell features chosen for the contrast cell features are all other sub-classes within the same broad class.
When searching for a specific gene, type the unique identifier into the "Filter by Gene Name, Gene Symbol or Entrez Gene ID" text box and either hit enter or click "Search". You can also narrow your search by selecting a cell features filter. Your search results will open in a heat map viewer.
When you do not have a gene marker to initiate your search, a differential search can be useful in that it will look for genes enhanced in the cell type you are interested in. To perform a differential search, you must select target and contrasting cell features by selecting from the matrix that opens when you click in either text box. The toggle switch to the right of the text boxes will exchange the Target and Contrast Cell Features.
Once you have selected your search criteria, clicking "Search" will open up your results in a heatmap viewer.
Clicking in the "Cell Features" text box will with the various cell features in this resource. A two-layer classification was assigned for each cell: the first "Class" indicates the broad class: Three classes of GABAergic neurons (Gad2), one glutamatergic class (Slc17a6), one non-neuronal class (Olig1), and one distinct class (Lars2_Kcnmb1). The second classification, "Subclass" identifies putative subclasses within each of these broad classes. The Classes and Subclasses were given names based on a set of marker genes that distinguish them. See the table below to understand how the classes and subclasses relate to each other.
Cell Classes and Subclasses
Non neuronal cells
The cell features are grouped by Class, Subclass, Region, Driver and Age. Each one of these features can be queried by clicking on the arrow to open a drop-down menu with each of the variables listed. Check the box(es) next to the variable you would like to search to filter by that feature.
Gad2_Chrna6, Gad2_Sepp1, Gad2_Syt4, Lars2_Kcnmb1, Olig1, Slc17a6
Choose a subclass from the #Cell Classes and Subclasses table
Mouse LGN - Core, Mouse LGN - Shell
Gad2-IRES-Cre, Slc17a6-IRES-Cre, Slc32a1-IRES-Cre, Snap25-IRES-Cre
Once you have conducted a gene or differential search or have selected one of the curated searches, your results will be loaded into a heatmap viewer. Once you have clicked on a data point in the heat map, metadata will be populated in the summary above the heatmap.
- Gene List: List of genes defined by the search criteria by gene symbol. When the list is a result of a differential search, each gene will be accompanied by both a "Fold Change" and "p-value". Selecting those column headers will toggle the sort of the gene list. When the list is the result of a correlative search each gene will be accompanied by a Pearson's correlation, r.
- Number of Genes: Number of genes that fit the search criteria.
- Column Headers: Clicking in this box will allow you to change the initial sort parameters of the column headers
- Classification: Indicates the class, subclass and region of the cell currently highlighted by the mouse
- Scroll Bars: Use the scroll bars to see the entire dataset.
- Gene Selection: Select genes by clicking on the checkbox next to the gene symbol in the gene list. View a heatmap with only selected genes by clicking "View Selected Genes". Genes will be available for viewing until you click "Clear Selections" or clear your cache.
- Filter Heatmap Function: To limit the amount of data displayed in the heat map use this function.
- Color Map: Use this function to change the way the z-scored data is displayed or to view the log2 FPKM data.
- Download: This link initiates download of the current heatmap data.
cell features can be used to sort the columns. Clicking the arrow in the box in the upper left-hand corner of the heatmap will open a list of groupings including the ones the user has created. To create a new grouping, click on "[Create new grouping...]", and a window will open allowing you to create a new grouping. Remember to save your selection.The column headers on the x-axis of the heat map are a feature that can be changed by the user. By default, the column headers are the class, subclass and region (in that order), but any of the
To remove a grouping, hover the mouse over the grouping and a garbage can will appear. Click on the can to remove the grouping.
To restrict the amount of data that are displayed in the heatmap, select the "Filter Heatmap" button below the heatmap. Filtering your heatmap is a two-step process: first, select the "..." box to restrict your features, making sure to save your selections, and then toggle the filter heatmap feature between "On" and "Off" by clicking the "Filter Heatmap" button.
Download Heatmap Data
Once you have found the data that you are interested in downloading and analyzing off-line, click on the "Download this data" link. The data will be downloaded as three .csv files; one with metadata for the rows (genes), one for the columns (cells) and a matrix containing the FPKM values.
To change the contrast of the heatmap display, click and drag the slider bars in the color scale below the heatmap. Clicking on the scale will open a window allowing the user to choose from several color scales or the log2 FPKM view.
Once you have found a gene of interest either by performing a gene or differential search, you can look for cell types that show a similar pattern of gene expression using the "Find Correlates" feature. This function will allow you to find cells with similar gene expression profiles as well as allowing you to find co-expressed genes within a single cell.
cell features before clicking "Search". All cell types with similar expression patterns of your seed gene will then be displayed in the heat map. If you filtered your search using cell features, your heatmap will only display those features. Turn off the "Filter Heatmap" function to see all the cell types.From the heatmap, click on a gene to load that "seed gene" into the search box. You have the option to select