Add new CountESS tutorial

[PlushZ]

Add a new Variant Analysis tutorial: Calculating CHEK2 variant effect scores from MAVE data with CountESS.

This tutorial introduces MAVE/DMS-style variant frequency data and shows how to run a saved CountESS workflow in Galaxy to calculate RAD53 Complementation Scores for CHEK2 coding SNVs. The tutorial uses a MaveDB-derived CHEK2 frequency summary and a CountESS .ini workflow, then explains how the calculated log-ratio scores relate to the original MaveDB score set.

[tflowers15]

Hi @PlushZ , Thank you for creating the new tutorial. It looks good. I do have a couple requests/suggestions.

Even though this is a short tutorial, could you please add a "Conclusion" section? This would be a good place to comment on some of the types of further analysis that the CountESS output can be used for.

Related to this, is there any simple next step(s) that you could include showing how the CountESS output (log ratios) can be used? Are there any plots that you can make to visualise the CountESS output?

[PlushZ]

Thanks @tflowers15 for the review and suggestions. I added a new Visualize the score distribution section with a short interpretation of what the plot shows. I also added a Conclusion section describing possible downstream uses of the CountESS output. Other changes include explanation of how preprocessed frequences were obtained, screenshots from the CountESS GUI for the details section, and added cancer context of CHEK2 to the intro. Let me know if you find something unclear.

[wm75]

dependent on galaxyproject/tools-iuc#7967

[tflowers15]

Hi @PlushZ , those are great additions to the tutorial and provide some nice context both on how the inputs are generated and what you can do with the results.

Sorry, I do again have another suggestion, but feel free to reject it if you think it is unnecessary. I think the "Compare with MaveDB" section could be clearer by adding some analysis steps to compare the RCS_this_SNV values in the MaveDB table and the table output from CountESS to demonstrate that the values are the same. A possible simple way to do this would be to use the Join tool to join the two tables on the mavedb accession column and then create a Scatterplot with ggplot2 plotting the MaveDB and CountESS RCS_this_SNV values on the x and y axes to show that all points lie along the 1:1 diagonal.

Otherwise, I think the tutorial is clear and easy to follow.

[PlushZ]

@tflowers15 thanks for your useful suggestions, I implemented them and think it improved the tutorial.

[PlushZ]

@tflowers15 Hi and thanks again for your useful suggestions that I implemented to the tutorial. galaxyproject/tools-iuc#7967 has been merged and tool installed, can we approve and merge this PR now?

[tflowers15]

@PlushZ, that's great! I have just tested the tutorial steps on Galaxy EU and made some minor changes to the tutorial as there were a couple tool input parameters that were named slightly differently than in Galaxy.

I also updated the input dataset icons for the tools using the chek2_rcs_scores dataset from the file to collection icon as these steps work with the collection containing the chek2_rcs_scores dataset as I don't think it is possible to have a tool run only on an individually specified dataset from inside a collection.

An alternative option would be to add a step before running Cut and running the Extract dataset from a collection by index or identifier tool on the CountESS output collection to extract the chek2_rcs_scores dataset and then not have to worry about collections in downstream analysis.

I have one additional request, but I wanted to ask you before adding it to the tutorial. Do you mind if I add my funding organisations to the funding section of the tutorial (The University of Melbourne, Melbourne Bioinformatics and Australian BioCommons)? The funding organisations are named as below in the ORGANISATIONS.yaml file.

• unimelb
• melbournebioinformatics
• AustralianBioCommons

If you approve of the changes I have made then I will approve and merge the pull request tomorrow (Thursday, July 30). I can also update the funding organisations, but I will only do that with your permission.

[PlushZ]

Hi @tflowers15 thanks for testing and adding your corrections, I forgot that I change some tool parameters names. Sure, you can add your funding organisations. All the rest looks good to me. Thank you!