When exploring data on Single Cell Portal, it’s often useful to get a quick impression of the genes that mark a cell type in those data. In this vein, Single Cell Portal offers exploratory differential expression (DE) results in for most studies with available raw count data. 

We’ve recently added some refinements that will make this feature more useful – read on to learn more!

Filters to isolate the most differentially expressed genes

If you’re exploring a study with differential expression results, you can select a specific cell type to compare to all other cell types. The results are summarized in a table that lists each gene’s log2fold change (log2FC) and adjusted p-value:

But this table displays the results for all genes in the study, and therefore might not highlight the genes that you’re most interested in. So we recently added filters to threshold these results based on p-value and log2FC.

These filters allow you to decide how you spotlight the genes that interest you most – whether that’s the genes with the largest expression difference between cell types, the smallest p-values, or some combination.

Explore differential expression results through dot plots

Isolating the genes that are most interesting to you only gets you part of the way toward understanding the data. Often it’s useful to look at plots to understand the patterns behind those statistics. To support this, you can now explore DE results through a dot plot.

Clicking the Dot plot button at the top right of the DE results table will render a dot plot that summarizes the expression patterns for all genes displayed in first page of the results table:

Viewing the results in a dot plot makes it easier to answer questions like, “was this gene expressed in all B cells or just a few?” or “was this gene only expressed in B cells, or also in other cell types?” Dot plots also make it easy to compare these patterns across genes – for example, did some genes have a similar pattern of expression across cell types?

You can also explore a single gene’s expression through a gene expression scatter plot, violin plot, or related genes ideogram. You can pull up these plots by clicking on the circle next to that gene in the results table.

Upload pre-computed differential expression results to your own study

Ultimately, you know your data – and how best to analyze it – better than we do. So, if you have differential expression results, we want to show those results to other researchers. If you own a study on SCP, you can now add pre-computed differential expression results to the study by navigating to the “Differential Expression” menu on the left-hand panel of the study's upload wizard: 

Use the interactive file format describer (top right) to learn more about the different accepted formats, or visit our documentation. Once you've uploaded your results, users who visit your study can explore them visually and through the results table described above.

Try it out!

You can explore these new features by checking out SCP1671 (the example study shown here). DE is available for studies that meet our DE analysis criteria. Check out our documentation for more details on how SCP computes differential expression results when they're not provided by the study owner.