Basic nextflow pipeline to run Proseg on a single sample
Pipeline parameters can be specified in nextflow.config
Key parameters include:
platform
-xenium
,cosmx
, ormerfish
outputDir
- where to store the output filesinputDir
- directory with the input file(s)transcriptsFile
- filename ininputDir
with transcript info. For Xenium this istranscripts.parquet
enforce_connectivity
- prevent cells from having disconnected voxelsignore_z_coord
- run in 2D mode
Right now proseg-workflow
only supports running Proseg on CosMx data with the --cosmx
flag. If there is interest
in supporting --cosmx-micron
and FOV stitching, please create an issue.
Make sure Nextflow and Proseg are installed.
A docker container with Proseg is provided easy deployment
To run the pipeline with docker:
nextflow run main.nf -with-docker [docker env name]
This workflow can be run on Cirro
On Cirro the workflow is launched with 16 CPUs and 64GB of memory on AWS. If the initial run crashes, nextflow will try to relaunch 3 times with increasing levels of memory (64GB, 128GB, 192GB).