run-umap.py

The python code to perform dimensionality reduction has the following package dependencies:

• numpy
• pandas
• umap
• sklearn

transcriptomoscope.R

The R code to plot spatial transcriptomics data using Voronoi tessellation has these dependencies:

• deldir
• sp
• rgeos
• alphahull
• RColorBrewer

How to use

Please refer to the command line help available with -h for both scripts, and to the following usage examples.

Usage examples

Simply plot the relative frequencies:

~/code/transcriptomoscope/transcriptomoscope.R counts*.tsv.gz

Plot the relative frequencies, allowing the plot coordinates to range from 0 to 1000 in both dimensions:

~/code/transcriptomoscope/transcriptomoscope.R counts*.tsv.gz -C 10000 -R 10000

This is useful when the coordinates of the spots are not in the standard range of 33 columns and 35 rows.

Perform dimensionality reduction with t-SNE and subsequently plot the results:

~/code/transcriptomoscope/run-umap.py --dim-red t-SNE  -o t-SNE_ counts*.tsv.gz
~/code/transcriptomoscope/transcriptomoscope.R --one t-SNE_joint_*txt -o t-SNE.pdf

Perform dimensionality reduction with UMAP and subsequently plot the results:

~/code/transcriptomoscope/run-umap.py --dim-red UMAP  -o umap_ counts*.tsv.gz
~/code/transcriptomoscope/transcriptomoscope.R --one umap_joint_*txt -o umap.pdf

Perform dimensionality reduction with UMAP and subsequently plot the results with non-default coordinates:

~/code/transcriptomoscope/run-umap.py --dim-red UMAP  -o umap_ counts*.tsv.gz
~/code/transcriptomoscope/transcriptomoscope.R --one umap_joint_*txt -o umap.pdf -C 10000 -R 10000