Differential Expression Viewer

View plugin for the ManiVault Studio framework for a basic comparison of genetic expression levels between selections (differential expression).

git clone https://github.com/ManiVaultStudio/DifferentialExpressionPlugin.git

This plugin captures variability between user-selected subsets by comparing the difference between expression averages per dimension. It computes the mean and median expression of all items (e.g. cells) in a selection and defines the differential expression as the difference of the means.

Information on more in-depth differential expression techniques can be found in many places, e.g. in Differential gene expression analysis pipelines and bioinformatic tools for the identification of specific biomarkers: A review (2024, 10.1016/j.csbj.2024.02.018).

How to use

1. Load a point data set, either via right-click in the data hierarchy and selecting View -> Differential Expression, by or opening an empty widget via the main toolbar with View -> Differential Expression View and then dragging-and-dropping the data into the small field that ask for a dataset.
2. Make two selection in the data, e.g. via a scatterplot view. Save each selection by clicking the respective buttons at the bottom of the view.
3. Click the button above the selection-setters to compute the differential expression.
4. The resulting DE computation will be listed in table form, with one row for each dimension of the data (listed in the ID column).
5. You can now sort the table along each column or use the search bar to filter the dimension names.

The "Min-max normalization" option scales both mean (and median) of each selection values with (selection_mean - global_min) / (global_max - global_min). The global_* values are computed for all data points, also those not selected.