broadinstitute/cellpainting_scripts

Helper scripts for processing images using CellProfiler

Helper scripts for processing images using CellProfiler

Note: The README on this page has now been moved into a handbook created using this repo.

The archived version of this repo lives here (private)

cellpainting_scripts's Issues

create_csvs_from_xml hangs if CSV already created

It gets stuck at an "Overwrite Y/N?" prompt and never proceeds.

Support alternate image directories when building CSVs

For example, if images need to be projected before being passed into the illum etc pipelines, we ideally like to upload them into an images_unprojected folder.

We right now can solve this in one of two ways- either initially uploading to images, then doing a whole bunch of file movement after the projection pipeline finishes to get the original images into images_unprojected and the newly projected ones into images; OR by uploading to images_unprojected, softlinking that to the bucket instead, hacking the create_csv_from_xml.sh and get_plate_dir_from_id.sh files as follows, then reverting after you finish the projection.

Neither is the end of the world, but being able to pass the subdir as an optional flag (with default images to avoid breaking behavior) would be a neat addon.

git diff
diff --git a/create_csv_from_xml.sh b/create_csv_from_xml.sh
index 4713768..6f54f14 100755
--- a/create_csv_from_xml.sh
+++ b/create_csv_from_xml.sh
@@ -71,7 +71,7 @@ pe2loaddata_dir=`readlink -m ${base_dir}/software/pe2loaddata`
 
 check_path exists ${pe2loaddata_dir}
 
-image_dir=`readlink -m ${base_dir}/images/$batch_id/"${plate_dir}"/Images/`
+image_dir=`readlink -m ${base_dir}/images_unprojected/$batch_id/"${plate_dir}"/Images/`
 
 check_path exists "${image_dir}"
 
diff --git a/get_plate_dir_from_id.sh b/get_plate_dir_from_id.sh
index 5f1badb..35bc080 100755
--- a/get_plate_dir_from_id.sh
+++ b/get_plate_dir_from_id.sh
@@ -22,7 +22,7 @@ plate_id=$2
 
 topdir=../..
 
-image_dir=`readlink -e ${topdir}/images/${batch_id}/`
+image_dir=`readlink -e ${topdir}/images_unprojected/${batch_id}/`