cellxgene-user-guide

Cell-x-gene user guide

cell-x-gene is a single-cell visualization platform developed by the Chan-Zuckerberg initative. It allows to explore single-cell RNA-seq (scRNA-seq) datasets in the web browser without any computational skills. This guide gives an overview of the most important features.

Table of Contents

Portal overview

picture 1

The portal consists of three main panels:

Additionally, there are two toolbars:

How to color by categorical variables (cell-type, patient, etc.)

In the left sidebar, click on the drop next to a variable. If you open another category (“patient” in the example), you’ll see colorbars indicating the distribution of the category.

picture 11

For some variables (cluster, cell-type) it makes sense to overlay the legend on the cell-type plot. To activate this feature, click on the “show labels” button in the toolbar:

picture 5

How to color by gene

In the right sidebar, enter a gene in the “Search” bar.

picture 6

This will add a histogram to the right sidebar. Click on the drop to color the single-cell plot by the gene expression:

picture 12

How to change the embedding

For some datasets, we provide multiple embeddings (usually, that is a UMAP with and without batch correction, respectively). You can choose the embedding from the “embedding toolbar” at the bottom left:

picture 8

How to analyse a subset of data (e.g. specific cell-type or patient)

There are different ways to select cells in cell-x-gene:

  1. By selecting categories in the left sidebar
  2. Using drag & drop on the plot
  3. By selecting a range in the right sidebar (drag & drop on a histogram)

Once you selected cells, you remove all other cells from the plot by clicking on the “Subset” button in the toolbar:

picture 9

Once you subset the dataset, all cell-counts, colorbars and histograms adapt to reflect the reduced dataset.

To reset the view to include all cells, use the “Undo subset” button. You can also undo a single step by using the “undo” button:

picture 10

How to perform differential gene expression analysis

With cell-x-gene it is possible to perform a very simple differential gene expression (DE)-analysis. It does not correct for any biases and only shows the top 10 most differentially expressed genes. This is great for a first glance at the data - if you need a more in-depth analysis, contact your Bioinformaticians.

Once you selected a group of cells, you can assign these cells to either group 1 or group 2 by clicking the buttons in the toolbar:

picture 15

Once you have both groups filled with cells, you can perform the DE-analysis by clicking on the “Differential expression” button:

picture 16

The 10 top DE genes will now appear in the right sidebar.