SCRAP-viz

Installing

To install (or upgrade), in your conda or python environment:

pip install --upgrade git+https://github.com/thomsonlab/scrap-viz.git

Running

Initializing a dataset

For each new dataset to analyze, you need to initialize a SCRAP-viz workspace for it. This workspace folder will automatically save labels you make and is where differential expression analyses will be exported to.

To initialize a dataset, from your command prompt, terminal, or conda prompt:

scrap-init-ws -w [path to the workspace you want to create] -s [path to your SDT file]

Workspaces can be initialized from H5, mtx, CSV, or SDT files using the related flag (-f, -m, -c, -s)

For example:

scrap-init-ws -w test_workspace -f ../data/filtered_feature_bc_matrix.h5

Will create a folder "test_workspace" in the current folder, and will initialize it to be a SCRAP-viz workspace based on the gene count data in the given H5 file.

Preprocessing a dataset

scrap-preprocess -w [path to the workspace] -p [what to call this preprocessing pipeline]

For example:

scrap-preprocess -w test_workspace -p my_preprocessing

Will process the gene counts in the "test_workspace" folder, and will name the preprocessing pipeline "my_preprocessing"

Running the GUI

scrap-viz -w [path to the workspace] -p [name of the preprocessing pipeline to use]

For example:

scrap-viz -w test_workspace -p my_preprocessing

Will launch a GUI server from the data in the "my_preprocessing" preprocessing pipeline