tutorial.md

layout	title	zenodo_link	level	questions	objectives	key_points	requirements	follow_up_training	time_estimation	contributors
tutorial_hands_on	Analyse HeLa fluorescence siRNA screen	https://zenodo.org/record/3362976	Intermediate	How do I analyze a HeLa fluorescence siRNA screen? How do I segment cell nuclei? How do I extract features from segmentations? How do I filter segmentations by morphological features? How do I apply a feature extraction workflow to a screen? How do I visualize feature extraction results?	How to segment cell nuclei in Galaxy. How to extract features from segmentations in Galaxy. How to filter segmentations by morphological features in Galaxy. How to extract features from an imaging screen in Galaxy. How to analyse extracted features from an imaging screen in Galaxy.	Galaxy workflows can be used to scale image analysis pipelines to whole screens. Segmented objects can be filtered using the **Filter segmentation** tool. Galaxy charts can be used to compare features extracted from screens showing cells with different treatments.	type topic_name tutorials internal imaging imaging-introduction	type topic_name tutorials internal statistics machinelearning	1H	thomaswollmann

Introduction

{:.no_toc}

This tutorial shows how to segment and extract features from cell nuclei Galaxy for image analysis. As example use case, this tutorial shows you how to compare the phenotypes of PLK1 threated cells in comparison to a control. The data used in this tutorial is available at Zenodo.

RNA interference (RNAi) is used in the example use case for silencing genes by way of mRNA degradation. Gene knockdown by this method is achieved by introducing small double-stranded interfering RNAs (siRNA) into the cytoplasm. Small interfering RNAs can originate from inside the cell or can be exogenously introduced into the cell. Once introduced into the cell, exogenous siRNAs are processed by the RNA-induced silencing complex (RISC).The siRNA is complementary to the target mRNA to be silenced, and the RISC uses the siRNA as a template for locating the target mRNA. After the RISC localizes to the target mRNA, the RNA is cleaved by a ribonuclease. RNAi is widely used as a laboratory technique for genetic functional analysis. RNAi in organisms such as C. elegans and Drosophila melanogaster provides a quick and inexpensive means of investigating gene function. Insights gained from experimental RNAi use may be useful in identifying potential therapeutic targets, drug development, or other applications. RNA interference is a very useful research tool, allowing investigators to carry out large genetic screens in an effort to identify targets for further research related to a particular pathway, drug, or phenotype.

The example used in this tutorial deals with PLK1 knocked down cells. PLK1 is an early trigger for G2/M transition. PLK1 supports the functional maturation of the centrosome in late G2/early prophase and establishment of the bipolar spindle. PLK1 is being studied as a target for cancer drugs. Many colon and lung cancers are caused by K-RAS mutations. These cancers are dependent on PLK1.

Agenda

In this tutorial, we will deal with:

1. TOC {:toc}

{: .agenda}

Getting data

The dataset required for this tutorial contains a screen of DAPI stained HeLa nuclei (more information). We will use a sample image from this dataset for training basic image processing skills in Galaxy.

{% icon hands_on %} Hands-on: Data upload

1. If you are logged in, create a new history for this tutorial

   {% include snippets/create_new_history.md %}

2. Import {% icon galaxy-upload %} the following dataset from Zenodo or from the data library (ask your instructor).

   • Important: Choose the type of data as zip.

   https://zenodo.org/record/3362976/files/B2.zip
   

   {% include snippets/import_via_link.md %} {% include snippets/import_from_data_library.md %}

3. Unzip file {% icon tool %} with the following parameters:

   • {% icon param-file %} "input_file": Zipped input file
   • "Extract single file": Single file
   • "Filepath": B2--W00026--P00001--Z00000--T00000--dapi.tif

4. Rename {% icon galaxy-pencil %} the dataset to testinput.tif

   {% include snippets/rename_dataset.md %}

5. Unzip file {% icon tool %} with the following parameters:

   • {% icon param-file %} "input_file": Zipped input file
   • "Extract single file": All files

6. Rename {% icon galaxy-pencil %} the resulting collection to control {% include snippets/rename_collection.md %}

7. Import {% icon galaxy-upload %} the following dataset from Zenodo or from the data library (ask your instructor).

   • Important: Choose the type of data as zip.

   https://zenodo.org/record/3362976/files/B3.zip
   

   {% include snippets/import_via_link.md %} {% include snippets/import_from_data_library.md %}

8. Unzip {% icon tool %} to extract the zipped screen:

   • {% icon param-file %} "input_file": Zipped input file
   • "Extract single file": All files

9. Rename {% icon galaxy-pencil %} the collection to PLK1

10. Upload {% icon galaxy-upload %} the following segmentation filter rules as a new pasted file (format: tabular):

    	area	eccentricity
    min	500	0.
    max	100000	0.5
    

    {% include snippets/create_new_file.md format="tabular" %}

11. Rename {% icon galaxy-pencil %} dataset to rules

    {% include snippets/rename_dataset.md %} {: .hands_on}

Create feature extraction workflow

First, we will create and test a workflow which extracts mean DAPI intensity, area, and major axis length of cell nuclei from an image.

{% icon hands_on %} Hands-on: Create feature extraction workflow

1. Filter Image {% icon tool %} with the following parameters to smooth the image:

   • "Image type": Gaussian Blur
   • "Radius/Sigma": 3
   • {% icon param-file %} "Source file": testinput.tif file

2. Auto Threshold {% icon tool %} with the following parameters to segment the image:

   • {% icon param-file %} "Source file": output of Filter image {% icon tool %}
   • "Threshold Algorithm": Otsu
   • "Dark Background": Yes

3. Split objects {% icon tool %} with the following parameters to split touching objects:

   • {% icon param-file %} "Source file": output of Auto Threshold {% icon tool %}
   • "Minimum distance between two objects.": 20

4. 2D Feature Extraction {% icon tool %} with the following parameters to extract features from the segmented objects:

   • {% icon param-file %} "Label file": output of Split objects {% icon tool %}
   • "Use original image to compute additional features.": No original image
   • "Select features to compute": Select features
   • "Available features":
     • {% icon param-check %} Add label id of label image
     • {% icon param-check %} Area
     • {% icon param-check %} Eccentricity
     • {% icon param-check %} Major Axis Length

5. Filter segmentation {% icon tool %} with the following parameters to filter the label map from 3. with the extracted features and a set of rules:

   • {% icon param-file %} "Source file": output of Split objects {% icon tool %}
   • {% icon param-file %} "Feature file": output of 2D Feature Extraction {% icon tool %}
   • {% icon param-file %} "Rules file": rules file

6. 2D Feature Extraction {% icon tool %} with the following parameters to extract features the final readout from the segmented objects:

   • {% icon param-file %} "Label file": output of Filter segmentation {% icon tool %}
   • "Use original image to compute additional features.": Use original image
   • {% icon param-file %} "Original image file": testinput.tif file
   • "Select features to compute": Select features
   • "Available features":
     • {% icon param-check %} Mean Intensity
     • {% icon param-check %} Area
     • {% icon param-check %} Major Axis Length

7. Now we can extract the workflow for batch processing

   • Name it "feature_extraction".

   {% include snippets/extract_workflow.md %}

8. Edit the workflow you just created

   • Name the inputs input image and filter rules.
   • Mark the results of steps 5 and 6 as outputs (by clicking on the asterisk next to the output name).

{: .hands_on}

The resulting workflow should look something like this:

Apply workflow to screen

Now we want to apply our extracted workflow to original data and merge the results. For this purpose, we create a workflow which uses the previously created workflow as subworkflow.

{% icon hands_on %} Hands-on: Create screen analysis workflow

1. Create a new workflow in the workflow editor.

   {% include snippets/create_new_workflow.md %}

2. Add a Input dataset collection node and name it input images

3. Add a Input dataset node and name it rules

4. Add the feature_extraction workflow as node.

   • {% icon param-file %} "input image": input images output of Input dataset collection {% icon tool %}
   • {% icon param-file %} "filter rules": rules output of Input dataset {% icon tool %}

5. Add a Collapse Collection {% icon tool %} node.

   • {% icon param-file %} "Collection of files to collapse into single dataset": output of feature_extraction workflow
   • "Keep one header line": Yes
   • "Append File name": No
   • Mark the tool output as workflow output

6. Save your workflow and name it analyze_screen {: .hands_on}

The resulting workflow should look something like this:

{% icon hands_on %} Hands-on: Run screen analysis workflow

1. Run the screen analysis workflow {% icon workflow %} on the control screen and the rules file

   {% include snippets/run_workflow.md %}

2. Run the screen analysis workflow {% icon workflow %} on the PLK1 screen and the rules file

{: .hands_on}

Plot feature extraction results

Finally, we want to plot the results for better interpretation.

{% icon hands_on %} Hands-on: Plot feature extraction results

1. Click on the Visualize this data {% icon galaxy-barchart %} icon of the Collapse Collection {% icon tool %} results.

2. Run Box plot with the following parameters:

   • "Provide a title": Screen features
   • "X-Axis label":
   • "Y-Axis label":
   • "1: Data series":
     • "Provide a label": Mean intensity
     • "Observations": Column 1
   • "2: Data series":
     • "Provide a label": Area
     • "Observations": Column 2
   • "3: Data series":
     • "Provide a label": Major axis length
     • "Observations": Column 3

   {% icon question %} Questions

   Plot the feature distribution of PLK1 and control. What differences do you observe between the screens?

   {% icon solution %} Solution

   The phenotype of PLK1 threated cells show a higher mean intensity and a shorter major axis in comparison to the control. {: .solution } {: .question} {: .hands_on}

One of the resulting plots should look something like this:

{: width="100%"}

Conclusion

{:.no_toc}

In this exercise you imported images into Galaxy, segmented cell nuclei, filtered segmentations by morphological features, extracted features from segmentations, scaled your workflow to a whole screen, and plotted the feature extraction results using Galaxy.