MTI_report

Generates an HTML single-sample report

Requirements

numpy
pandas
anndata
scanpy
squidpy
matplotlib
plotly
sklearn
markdown

Or use env.yml to create conda environment

Basic Usage

./sample_report.py -f my_data.h5ad

Input data should contain a phenotype column in adata.obs

usage: sample_report.py [-h] -f FILE [-c COLUMN] [--removeMarkers REMOVEMARKERS [REMOVEMARKERS ...]] [--nuclearMarker NUCLEARMARKER] [--resolution RESOLUTION] [--leidenRange LEIDENRANGE [LEIDENRANGE ...]] [--radius RADIUS]

optional arguments:

-h, --help show this help message and exit

-f FILE, --file FILE anndata h5ad cell feature table

-c COLUMN, --column COLUMN column name for phenotypes, default is 'phenotype'

--removeMarkers REMOVEMARKERS [REMOVEMARKERS ...] Patterns to remove markers by. Example: 'DAPI' will remove 'DAPI_1','DAPI_2', etc. Markers are removed from adata.X but maintained in raw. default is 'DAPI' 'AF'

--nuclearMarker NUCLEARMARKER Name of nuclear marker in every cycle for QC purposes, default is 'DAPI'

--resolution RESOLUTION Image resolution in microns/px for density calculation (default is 0.65 micron/px)

--leidenRange LEIDENRANGE [LEIDENRANGE ...] list of resolutions to test (default is just [0.5,0.6])

--radius RADIUS Radius (pixels) for neighborhood search (default is 30px)

Output

In the current working directory, sample_report.py will produce:

- figures (dir)
- output anndata file (basename_analyzed.h5ad)
- HMTL report with selected figures (basename_report.html)

Between runs, rename or delete figures dir