V. cholerae O1/O139 mlst scheme selection bugfix

This PR closes #58

🗑️ This dev branch should be deleted after merging to main.

:brain: Aim, Context and Functionality

Vibrio cholerae has two MLST schemes, one for O1 or O139 serogroups, and one for non-O1/non-O139 serogroups. The latter is always being used regardless of serogroup, because mlst's software auto scheme detection feature seems to always choose the non-O1 & non-O139 scheme. mlst also has a few default schemes that are excluded by default and vcholerae_2 scheme is one of them, so perhaps that is why vcholerae is default.

The mlst command-line tool calls these schemes as such:

Non-O1 & Non-O139 serogroup samples : vcholerae
O1 & O139 serogroup samples, vcholerae_2

By default, the mlst auto-scheme detection chooses the non-O1 & O139 scheme so this is a way of automatically forcing mlst to use vcholerae_2 for samples detected as O1 or O139 via the SRST2 task.

More info on the 2 mlst schemes can be found here: https://pubmlst.org/organisms/vibrio-cholerae

:hammer_and_wrench: Impacted Workflows/Tasks & Changes Being Made

This will affect the behavior of the workflow(s) even if users don’t change any workflow inputs relative to the last version : No, but the exception is if they are analyzing V. cholerae samples that are O1 or O139 serogroups. The mlst results will change if analyzed w/ a previous version of TheiaProk wf

Running this workflow on different occasions could result in different results, e.g. due to use of a live database, "latest" docker image, or stochastic data processing : No

:clipboard: Workflow/Task Step Changes

🔄 Data Processing

Docker/software or software versions changed: No

Databases or database versions changed: No

Data processing/commands changed: No

File processing changed: No

Compute resources changed: No

➡️ Inputs

ts_mlst.scheme has been exposed as an optional input for the TheiaProk_Illumina_PE workflow

⬅️ Outputs

:test_tube: Testing

Test Dataset

Commandline Testing with MiniWDL or Cromwell (optional)

Terra Testing

Suggested Scenarios for Reviewer to Test

Theiagen Version Release Testing (optional)

:microscope: Final Developer Checklist

[x] The workflow/task has been tested locally and results, including file contents, are as anticipated
[x] The workflow/task has been tested on Terra and results, including file contents, are as anticipated
[x] The CI/CD has been adjusted and tests are passing (to be completed by Theiagen developer)
[x] Code changes follow the style guide

🎯 Reviewer Checklist

[ ] All impacted workflows/tasks have been tested on Terra with a different dataset than used for development
[ ] All reviewer-suggested scenarios have been tested and any additional
[ ] All changed results have been confirmed to be accurate
[ ] All workflows/tasks impacted by change/s have been tested using a standard validation dataset to ensure no unintended change of functionality
[ ] All code adheres to the style guide
[ ] MD5 sums have been updated
[ ] The PR author has addressed all comments

🗂️ Associated Documentation (to be completed by Theiagen developer)

[x] Relevant documentation on the Public Health Resources "PHB Main" has been updated. No need to update docs unless we want to add a caveat/note about the auto-scheme selection for Vcholerae O1 and O139 samples
[x] Workflow diagrams have been updated to reflect changes. No need to update wf diagram

TODO: need to update theiaprok ILMN SE, ONT, and FASTA workflows to get mlst scheme input from merlin_magic subworkflow.

~~Testing that this works for ILMN PE in Terra now~~

Tests launched & comments regarding them:

TheiaProk_Illumina_PE; PHE Vibrio cholerae dataset, mix of O1's, O139's, and non O1s & O139s; did not provide any optional input for mlst_scheme: https://app.terra.bio/#workspaces/theiagen-validations/curtis-sandbox-theiagen-validations/job_history/464f6fef-f9f0-4a36-929b-265cf8c2567f
- ran as expected ✅ , used vcholerae_2 scheme for those designated as O1/O139 from SRST2 task and used vcholerae for all other Vcholerae samples
TheiaProk_Illumina_PE; A. baumanii samples; did not provide mlst_scheme as input: https://app.terra.bio/#workspaces/theiagen-validations/curtis-sandbox-theiagen-validations/job_history/17e64f6a-97b5-441f-9a0e-57d4ac764066
- ran as expected, and used the 2 single quotes "''" as input for mlst_scheme to trigger the mlst scheme auto-selection feature
TheiaProk_Illumina_PE; 2 Shigella samples; DID provide mlst_scheme as input (used "ecoli_achtman_4"): https://app.terra.bio/#workspaces/theiagen-validations/curtis-sandbox-theiagen-validations/job_history/942e9f52-4eaf-4446-9b0c-34f76fae8c25
- ran as expected, used the higher-priority mlst_scheme provided input

TODO:

[x] functional test of TheiaProk_FASTA (at least 1 sample, does not need to be Vcholerae)
[x] functional test of TheiaProk_ONT (at least 1 sample, does not need to be Vcholerae)
[x] functional test of TheiaProk_Illumina_SE (at least 1 sample, does not need to be Vcholerae)

Functional test results:

TheiaProk_FASTA; mixture of cholerae and non-cholerae (don't know serogroup & these are assemblies anyways so it doesn't matter): https://app.terra.bio/#workspaces/theiagen-validations/curtis-sandbox-theiagen-validations/job_history/42ca5c94-e306-4573-8aaa-56cbec6ad822
- did not change input mlst_scheme variable, left blank
- ran as expected, and used the 2 single quotes "''" as input for mlst_scheme to trigger the mlst scheme auto-selection feature
TheiaProk_ONT; a few S. pyogenes samples: https://app.terra.bio/#workspaces/theiagen-validations/curtis-sandbox-theiagen-validations/job_history/e78e16d8-a072-4c0e-bc9a-d39ae39ca41f
- did not change input mlst_scheme variable, left blank
- ran as expected, and used the 2 single quotes "''" as input for mlst_scheme to trigger the mlst scheme auto-selection feature
TheiaProk_Illumina_SE: https://app.terra.bio/#workspaces/theiagen-validations/curtis-sandbox-theiagen-validations/job_history/4065c6a4-c94f-4307-8535-3ff9c6c43300
- did not change input mlst_scheme variable, left blank
- ran as expected, and used the 2 single quotes "''" as input for mlst_scheme to trigger the mlst scheme auto-selection feature

One last test to show that user-defined input is priority in the select_first statement in merlin_magic subworkflow.

I used "senterica" as the input mlst_scheme variable for a few E. coli and the mlst task used senterica as the scheme:

https://app.terra.bio/#workspaces/theiagen-validations/curtis-sandbox-theiagen-validations/job_history/ddd76e8c-fc7e-467d-a3e5-307cec43020c

I have looked deeper into the literature given Curtis' observation that no ST was assigned to the O1 and O139 samples with the vcholerae_2 scheme.

Whilst PubMLST lists the Kanampalliwar & Singh scheme (vcholerae_2)published in 2020 as the O1 and O139 scheme, it doesn't seem to be used widely (only 5 citations to the paper) and only 400 PubMLST submissions to the scheme have ever been made, all on one date, presumably when the scheme was first published. The paper publishing the scheme. described that "there is no proper MLST database for typing V. cholerae O1 and O139 strains, except for non-O1, non-O139 strains (Octavia et al., 2013)." hence them developing their scheme.
The Octavia et al. scheme (vcholerae) published in 2013 is listed as the non-O1, non-O139 on PubMLST and the title of the paper publishing this scheme mentions only these serotypes, non-O1 and non-O139 . On reading the paper, the scheme was developed based on housekeeping genes identified in previous work looking at more diverse V. cholerae (including O1). Octavia et al. had published the MLST scheme in a paper that was also describing non-O1 and non-O139 isolates, but I can't see any references to this MLST scheme only being for non-O1/non-O139 other than on PubMLST and in the Kanampalliwar & Singh paper. The Octavia et al. scheme has been used widely, with 85 citations on Google scholar. These seem to mostly be people using the MLST scheme, and they use it for O1 and O139. It also has 1800 submissions to PubMLST.

Though PubMLST lists the vcholerae_2 scheme as the O1 and O139 scheme, it may be more useful for the public health community to use the vcholerae scheme as they are more likely to identify a named ST for their V. cholerae, including O1 and O139 isolates, and the use of this scheme doesn't seem to be wrong- in fact, there seems to be a consensus to preferentially using this.

Thank you @emmadoughty for the detective work 🕵🏻‍♀️

I agree with your assessment that the community is using vcholerae scheme for all isolates, regardless of serotype. I think we should close off this PR and not incorporate these changes into the main codebase to keep in line with the community standards and expectations.

theiagen / public_health_bioinformatics