sciviewer 0.1

Tool for interactive exploration of 2D embeddings in Jupyter

Single Cell Interactive Viewer

This is an interactive viewer for Jupyter notebooks that allows users exploring an embedding of single-cell RNA-seq data. The user can select directions in the 2D embedding plane (e.g.: UMAP or tSNE), and the viewer will calculate the genes with the highest correlation along those directions. In addition, you can select a group of cells and find the genes that are most differentially expressed.

Installation

• Prepare conda environment containing dependencies for py5: conda env create -n sciviewer -f https://raw.githubusercontent.com/colabobio/sciviewer/master/sciviewer-env.yml
• Activate the environment created above: conda activate sciviewer
• Alternatively, if you want to append the needed dependencies to an existing conda environment, you can do the following:
  • conda env update -n your_existing_environment -f https://raw.githubusercontent.com/colabobio/sciviewer/master/sciviewer-env.yml
  • conda activate your_existing_environment
• Install the sciviewer package: python -m pip install --index-url https://test.pypi.org/simple/ --no-deps sciviewer
• To uninstall: pip uninstall sciviewer
• To load the module from a notebook wihtout installing (good for debugging), add the following imports to the notebook:

 import sys
 sys.path.insert(0, '../sciviewer/')
 from sciviewer import SCIViewer

• Note, the above won't work if you don't remove the relative imports from the local .py files.

Jupyter notebook

This Jupyter notebook demonstrates the use of py5, a version of Processing for Python, to create an interactive viewer of RNA-seq embedding data. Find it under the notebook folder.

Processing's drawing API and engine can be run from a Jupyter notebook to enable users explore their datasets and then bring the results of their exploration directly into the notebook for further analysis:

The datasets can be quite large, as seen in the next image: