A pipeline for pathology slide analysis: tile and detect tissue
- Unix-like operating system
- Java 11
-
Install nextflow:
curl -s https://get.nextflow.io | bash
-
Launch the pipeline execution using docker
./nextflow run msk-mind/tiling-nf --samples_csv samples.csv
samples.csv
is a csv file which has aurl
(specify with--slide_url_column
) column that has the path to the slide -
When the execution completes,
meta.csv
will be in theresults
(specify with--outdir
) directory
-
Create an Azure storage account and batch account.
-
Modify the nextflow configuration files as necessary (see https://www.nextflow.io/docs/edge/azure.html)
-
Set the necessary secrets using
nextflow secrets set
. At a minimum setAZURE_BATCH_KEY
andAZURE_STORAGE_KEY
. -
Launch nextflow:
nextflow -Dcom.amazonaws.sdk.disableCertChecking=true run main.nf -bucket-dir {azure_bucket_dir} -profile standard,cloud
where
{azure_bucket_dir}
is an azure path likeaz://test/nftest
. -
When the execution completes,
meta.csv
will be in theresults
directory