Add histological staining area quantification workflow by dianichj · Pull Request #1206 · galaxyproject/iwc

dianichj · 2026-04-09T10:29:38Z

This workflow quantifies the area of histological staining in brightfield
histological images using colour deconvolution and automated thresholding.
Compatible with Masson's Trichrome (MT) and Immunohistochemistry (IHC).

Published workflow: https://usegalaxy.eu/u/dianitachj24/w/histological-staining-area-quantification

Status

Draft PR - test data and .dockstore.yml will be added before final merge.

test-data to Zenodo (files > 100 KB)
CHANGELOG.md
.dockstore.yml

cc: @mvdbeek @afgane

Copilot

Pull request overview

Note

Copilot was unable to run its full agentic suite in this review.

Adds a Galaxy workflow to quantify histological staining area in brightfield images using colour deconvolution + thresholding, along with accompanying documentation and a workflow test definition.

Changes:

Added the “Histological Staining Area Quantification” Galaxy workflow (.ga) with steps for deconvolution, channel selection, thresholding, feature extraction, and result collation.
Added a workflow test YAML describing a multi-sample input collection and expected outputs.
Added a README describing inputs/outputs and compatible stainings.

Reviewed changes

Copilot reviewed 3 out of 3 changed files in this pull request and generated 10 comments.

File	Description
workflows/imaging/histological-staining-area-quantification/histological-staining-area-quantification.ga	Introduces the full staining-quantification workflow and output collation steps.
workflows/imaging/histological-staining-area-quantification/histological-staining-area-quantification-test.yml	Adds a test definition for running the workflow on a collection of ROI images and checking outputs.
workflows/imaging/histological-staining-area-quantification/README.md	Documents workflow purpose, inputs/outputs, and compatible stain/channel assumptions.

Copilot · 2026-04-09T12:48:50Z

+            "workflow_outputs": []
+        },
+        "11": {
+            "annotation": "Applies a threshold to the selected stain channel to segment stained areas for quantification.\nTool only process tiff files with 1 channel. ",


The PR description (and workflow narrative) says the workflow uses Li thresholding, but the configured threshold method here is Otsu (method_id: otsu). This changes the core segmentation behavior. Either switch the workflow configuration to Li (if intended) or update the PR description/readme/annotations to reflect Otsu to keep the published method consistent.

Copilot · 2026-04-09T12:48:50Z

+                "owner": "imgteam",
+                "tool_shed": "toolshed.g2.bx.psu.edu"
+            },
+            "tool_state": "{\"input\": {\"__class__\": \"RuntimeValue\"}, \"invert_output\": false, \"th_method\": {\"method_id\": \"otsu\", \"__current_case__\": 1, \"offset\": \"0.0\"}, \"__page__\": 0, \"__rerun_remap_job_id__\": null}",


The PR description (and workflow narrative) says the workflow uses Li thresholding, but the configured threshold method here is Otsu (method_id: otsu). This changes the core segmentation behavior. Either switch the workflow configuration to Li (if intended) or update the PR description/readme/annotations to reflect Otsu to keep the published method consistent.

Copilot · 2026-04-09T12:48:51Z

+            "workflow_outputs": [
+                {
+                    "label": "Tabular File: Staining Feature Results",
+                    "output_name": "out_file1",
+                    "uuid": "1c3f7fd0-bad7-4891-9f95-f1346e44c20f"
+                }
+            ]


The workflow computes percent_area in step 27, but step 27 is not exposed as a workflow output (workflow_outputs is empty). This makes the documented percent_area result unavailable to workflow consumers. Mark step 27 (or the appropriate downstream dataset that includes percent_area) as a workflow output, and adjust the test expectations accordingly.

Copilot · 2026-04-09T12:48:51Z

+            "type": "tool",
+            "uuid": "ae102451-2b3c-4ede-89db-a18e52b7bb2a",
+            "when": null,
+            "workflow_outputs": []


The workflow computes percent_area in step 27, but step 27 is not exposed as a workflow output (workflow_outputs is empty). This makes the documented percent_area result unavailable to workflow consumers. Mark step 27 (or the appropriate downstream dataset that includes percent_area) as a workflow output, and adjust the test expectations accordingly.

Suggested change

"workflow_outputs": []

"workflow_outputs": [

{

"label": "percent_area",

"output_name": "out_file1",

"uuid": "2d53a6f8-0a7e-4c55-9f52-cfdff8e6267e"

}

]

Copilot · 2026-04-09T12:48:51Z

+| Staining | Target channel |
+|----------|---------------|
+| Masson's Trichrome (MT) | Channel 2 |
+| Immunohistochemistry (IHC) | Channel 2 (DAB) |


The markdown table is malformed (uses || at line starts), which may render incorrectly. Use standard markdown table syntax with single leading/trailing | per row so the table renders reliably.

Copilot · 2026-04-09T12:48:52Z

+            "label": "Extract",
+            "name": "Select last",


The workflow selects the last row from each per-sample feature table before collapsing. For binary masks, feature extraction commonly yields multiple labels (e.g., background=0 and stained=1). Relying on row order is brittle and can select the wrong label if ordering changes. Prefer filtering explicitly on the label column (e.g., keep only label == 1) before collapsing, or configure the feature-extraction step to output only the target label if the tool supports it.

Copilot · 2026-04-09T12:48:52Z

+            },
+            "post_job_actions": {},
+            "tool_id": "Show tail1",
+            "tool_state": "{\"header\": false, \"input\": {\"__class__\": \"ConnectedValue\"}, \"lineNum\": \"1\", \"__page__\": 0, \"__rerun_remap_job_id__\": null}",


The workflow selects the last row from each per-sample feature table before collapsing. For binary masks, feature extraction commonly yields multiple labels (e.g., background=0 and stained=1). Relying on row order is brittle and can select the wrong label if ordering changes. Prefer filtering explicitly on the label column (e.g., keep only label == 1) before collapsing, or configure the feature-extraction step to output only the target label if the tool supports it.

Copilot · 2026-04-09T12:48:52Z

+              path: 'test-data/Collection: Individual Deconvolved Channels_1.tiff.tiff'
+            2.tiff:
+              path: 'test-data/Collection: Individual Deconvolved Channels_2.tiff.tiff'
+            3.tiff:
+              path: 'test-data/Collection: Individual Deconvolved Channels_3.tiff.tiff'
+        control_2:
+          elements:
+            1.tiff:
+              path: 'test-data/Collection: Individual Deconvolved Channels_1.tiff.tiff'
+            2.tiff:
+              path: 'test-data/Collection: Individual Deconvolved Channels_2.tiff.tiff'
+            3.tiff:
+              path: 'test-data/Collection: Individual Deconvolved Channels_3.tiff.tiff'
+        control_3:
+          elements:
+            1.tiff:
+              path: 'test-data/Collection: Individual Deconvolved Channels_1.tiff.tiff'
+            2.tiff:
+              path: 'test-data/Collection: Individual Deconvolved Channels_2.tiff.tiff'
+            3.tiff:
+              path: 'test-data/Collection: Individual Deconvolved Channels_3.tiff.tiff'
+        treatment_1:
+          elements:
+            1.tiff:
+              path: 'test-data/Collection: Individual Deconvolved Channels_1.tiff.tiff'
+            2.tiff:
+              path: 'test-data/Collection: Individual Deconvolved Channels_2.tiff.tiff'
+            3.tiff:
+              path: 'test-data/Collection: Individual Deconvolved Channels_3.tiff.tiff'
+        treatment_2:
+          elements:
+            1.tiff:
+              path: 'test-data/Collection: Individual Deconvolved Channels_1.tiff.tiff'
+            2.tiff:
+              path: 'test-data/Collection: Individual Deconvolved Channels_2.tiff.tiff'
+            3.tiff:
+              path: 'test-data/Collection: Individual Deconvolved Channels_3.tiff.tiff'
+        treatment_3:
+          elements:
+            1.tiff:
+              path: 'test-data/Collection: Individual Deconvolved Channels_1.tiff.tiff'
+            2.tiff:
+              path: 'test-data/Collection: Individual Deconvolved Channels_2.tiff.tiff'
+            3.tiff:
+              path: 'test-data/Collection: Individual Deconvolved Channels_3.tiff.tiff'


Several element_tests reuse the same expected output files across different input samples (e.g., control_2 points to the same deconvolved-channel golden files as control_1). Unless the inputs are identical, this will either fail once real test data is added or will mask regressions if comparisons are unintentionally satisfied. Recommend providing per-sample golden outputs (or switching to more robust assertions for images/tabular outputs if exact matches are not feasible).

Suggested change

path: 'test-data/Collection: Individual Deconvolved Channels_1.tiff.tiff'

2.tiff:

path: 'test-data/Collection: Individual Deconvolved Channels_2.tiff.tiff'

3.tiff:

path: 'test-data/Collection: Individual Deconvolved Channels_3.tiff.tiff'

control_2:

elements:

1.tiff:

path: 'test-data/Collection: Individual Deconvolved Channels_1.tiff.tiff'

2.tiff:

path: 'test-data/Collection: Individual Deconvolved Channels_2.tiff.tiff'

3.tiff:

path: 'test-data/Collection: Individual Deconvolved Channels_3.tiff.tiff'

control_3:

elements:

1.tiff:

path: 'test-data/Collection: Individual Deconvolved Channels_1.tiff.tiff'

2.tiff:

path: 'test-data/Collection: Individual Deconvolved Channels_2.tiff.tiff'

3.tiff:

path: 'test-data/Collection: Individual Deconvolved Channels_3.tiff.tiff'

treatment_1:

elements:

1.tiff:

path: 'test-data/Collection: Individual Deconvolved Channels_1.tiff.tiff'

2.tiff:

path: 'test-data/Collection: Individual Deconvolved Channels_2.tiff.tiff'

3.tiff:

path: 'test-data/Collection: Individual Deconvolved Channels_3.tiff.tiff'

treatment_2:

elements:

1.tiff:

path: 'test-data/Collection: Individual Deconvolved Channels_1.tiff.tiff'

2.tiff:

path: 'test-data/Collection: Individual Deconvolved Channels_2.tiff.tiff'

3.tiff:

path: 'test-data/Collection: Individual Deconvolved Channels_3.tiff.tiff'

treatment_3:

elements:

1.tiff:

path: 'test-data/Collection: Individual Deconvolved Channels_1.tiff.tiff'

2.tiff:

path: 'test-data/Collection: Individual Deconvolved Channels_2.tiff.tiff'

3.tiff:

path: 'test-data/Collection: Individual Deconvolved Channels_3.tiff.tiff'

path: 'test-data/Collection: Individual Deconvolved Channels_control_1_1.tiff.tiff'

2.tiff:

path: 'test-data/Collection: Individual Deconvolved Channels_control_1_2.tiff.tiff'

3.tiff:

path: 'test-data/Collection: Individual Deconvolved Channels_control_1_3.tiff.tiff'

control_2:

elements:

1.tiff:

path: 'test-data/Collection: Individual Deconvolved Channels_control_2_1.tiff.tiff'

2.tiff:

path: 'test-data/Collection: Individual Deconvolved Channels_control_2_2.tiff.tiff'

3.tiff:

path: 'test-data/Collection: Individual Deconvolved Channels_control_2_3.tiff.tiff'

control_3:

elements:

1.tiff:

path: 'test-data/Collection: Individual Deconvolved Channels_control_3_1.tiff.tiff'

2.tiff:

path: 'test-data/Collection: Individual Deconvolved Channels_control_3_2.tiff.tiff'

3.tiff:

path: 'test-data/Collection: Individual Deconvolved Channels_control_3_3.tiff.tiff'

treatment_1:

elements:

1.tiff:

path: 'test-data/Collection: Individual Deconvolved Channels_treatment_1_1.tiff.tiff'

2.tiff:

path: 'test-data/Collection: Individual Deconvolved Channels_treatment_1_2.tiff.tiff'

3.tiff:

path: 'test-data/Collection: Individual Deconvolved Channels_treatment_1_3.tiff.tiff'

treatment_2:

elements:

1.tiff:

path: 'test-data/Collection: Individual Deconvolved Channels_treatment_2_1.tiff.tiff'

2.tiff:

path: 'test-data/Collection: Individual Deconvolved Channels_treatment_2_2.tiff.tiff'

3.tiff:

path: 'test-data/Collection: Individual Deconvolved Channels_treatment_2_3.tiff.tiff'

treatment_3:

elements:

1.tiff:

path: 'test-data/Collection: Individual Deconvolved Channels_treatment_3_1.tiff.tiff'

2.tiff:

path: 'test-data/Collection: Individual Deconvolved Channels_treatment_3_2.tiff.tiff'

3.tiff:

path: 'test-data/Collection: Individual Deconvolved Channels_treatment_3_3.tiff.tiff'

Copilot · 2026-04-09T12:48:53Z

+            "workflow_outputs": []
+        },
+        "1": {
+            "annotation": "Add Sample ID column name .",


Fix minor grammar/typos in step annotations to improve readability (e.g., remove the extra space before the period in "name ." and change "Tool only process" to "Tool only processes").

Suggested change

"annotation": "Add Sample ID column name .",

"annotation": "Add Sample ID column name.",

Copilot · 2026-04-09T12:48:53Z

+            "workflow_outputs": []
+        },
+        "11": {
+            "annotation": "Applies a threshold to the selected stain channel to segment stained areas for quantification.\nTool only process tiff files with 1 channel. ",


Fix minor grammar/typos in step annotations to improve readability (e.g., remove the extra space before the period in "name ." and change "Tool only process" to "Tool only processes").

Suggested change

"annotation": "Applies a threshold to the selected stain channel to segment stained areas for quantification.\nTool only process tiff files with 1 channel. ",

"annotation": "Applies a threshold to the selected stain channel to segment stained areas for quantification.\nTool only processes TIFF files with 1 channel.",

dianichj added 2 commits April 9, 2026 12:14

Add histological staining area quantification workflow

567e5c2

Fix ORCID URI and add README

36f7879

mvdbeek requested a review from Copilot April 9, 2026 12:40

Copilot AI reviewed Apr 9, 2026

View reviewed changes

Copilot started reviewing on behalf of mvdbeek April 9, 2026 12:57 View session

add report template to workflow

ae7d395

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Add histological staining area quantification workflow#1206

Add histological staining area quantification workflow#1206
dianichj wants to merge 3 commits intogalaxyproject:mainfrom
dianichj:histological-staining-quantification-workflow

dianichj commented Apr 9, 2026

Uh oh!

Copilot AI left a comment

Uh oh!

Copilot AI Apr 9, 2026

Uh oh!

Copilot AI Apr 9, 2026

Uh oh!

Copilot AI Apr 9, 2026

Uh oh!

Copilot AI Apr 9, 2026

Uh oh!

Copilot AI Apr 9, 2026

Uh oh!

Copilot AI Apr 9, 2026

Uh oh!

Copilot AI Apr 9, 2026

Uh oh!

Copilot AI Apr 9, 2026

Uh oh!

Copilot AI Apr 9, 2026

Uh oh!

Copilot AI Apr 9, 2026

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

3 participants

	"annotation": "Add Sample ID column name .",
	"annotation": "Add Sample ID column name.",

	"annotation": "Applies a threshold to the selected stain channel to segment stained areas for quantification.\nTool only process tiff files with 1 channel. ",
	"annotation": "Applies a threshold to the selected stain channel to segment stained areas for quantification.\nTool only processes TIFF files with 1 channel.",

Conversation

dianichj commented Apr 9, 2026

Status

Uh oh!

Copilot AI left a comment

Choose a reason for hiding this comment

Pull request overview

Reviewed changes

Uh oh!

Copilot AI Apr 9, 2026

Choose a reason for hiding this comment

Uh oh!

Copilot AI Apr 9, 2026

Choose a reason for hiding this comment

Uh oh!

Copilot AI Apr 9, 2026

Choose a reason for hiding this comment

Uh oh!

Copilot AI Apr 9, 2026

Choose a reason for hiding this comment

Uh oh!

Copilot AI Apr 9, 2026

Choose a reason for hiding this comment

Uh oh!

Copilot AI Apr 9, 2026

Choose a reason for hiding this comment

Uh oh!

Copilot AI Apr 9, 2026

Choose a reason for hiding this comment

Uh oh!

Copilot AI Apr 9, 2026

Choose a reason for hiding this comment

Uh oh!

Copilot AI Apr 9, 2026

Choose a reason for hiding this comment

Uh oh!

Copilot AI Apr 9, 2026

Choose a reason for hiding this comment

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

3 participants