ClusterExploration

This extension allows to inspect clustering/classification outputs in image and linked embedding space.

Installation

For End Users

Download latest release and execute it. This will start a local server on your machine. Open the browser and navigate to: http://mts-lsp-l06208:8000/

For Developers

1. Clone locally from Github: git clone -b gating git@github.com:labsyspharm/minerva_anaysis.git
   • or, switch to branch: git switch embedding
2. Run from readme.md (no new env updates)

Data

Beyond the standard imaging data, we expect the embedding as a 3 or 4 column csv in the order cellId, x, y, clusterId(optional), and optionally, a file in the "Cell Type File" field that is a csv with 2 columns, clusterId, clusterName which maps the id to a string describing the cluster.

Tips

• Hold shift to lasso select from either of the two spaces and drag.
• Click on the colored square in the legend to select all cells of that type in both spaces
• Nuclei Mask is better than cell mask as it obscures less.