This PR updates the staphb/artic Docker image version from 1.2.4-1.11.1 to 1.2.4-1.12.0 in the Medaka task in the ONT assembly workflow to handle the most recent basecaller/medaka models. The medaka model can now be auto-detected from the FASTQ. I decided to keep medaka_model as an optional input because some FASTQ may not have model information in the headers (for example downloaded from SRA).
Workflow Changes ✅
Upstream Effects
None
Input Changes
medaka_model is optional and should generally be blank to allow auto detection (avoids using the wrong model).
Output Changes
The assembly version capture file has a new row recording medaka_model version.
Downstream Effects
[ ] @danpolanco: Update affected BigQuery data transfers
Testing 🛠️
test_cov_2205_grid (new basecaller version- r1041_e82_400bps_hac_v4.3.0)
cov_2205_grid (old basecaller version - r1041_e82_400bps_hac_v4.2.0)
Test(s) performed:
Ran test_cov_2205_grid with and without model auto-detection, checked outputs are the same, did not crash.
Ran cov_2205_grid with model auto-detection and compared summary results to previous results with TheiaValidate.
Developer Checklist 👷♀️
[x] Prior to development, issues were discussed with the bioinformatics team members and approved
[x] Code has been refactored to sufficiently address the issues this pull request closes
[x] Testing was performed and the results from testing match the expected results
[x] All code changes match our style guide
[x] README has been updated to reflect changes
[x] Workflow diagrams in READMEs have been updated to reflect changes
Reviewer Checklist 🔍
[x] Met with developer to review all changes and testing performed
[x] Code refactoring sufficiently address the issue(s) that this pull request closes
[x] All code meets style guide critera
[x] Confirm testing was sufficient (e.g. correct dataset was used for testing, results match the expected results). If not, the developer will perform additional testing which should be documented in the testing section above.
[x] New version release number has been decided upon (if applicable)
This PR closes #74
Aim, context, and functionality 🎯
This PR updates the
staphb/articDocker image version from1.2.4-1.11.1to1.2.4-1.12.0in theMedakatask in the ONT assembly workflow to handle the most recent basecaller/medaka models. The medaka model can now be auto-detected from the FASTQ. I decided to keepmedaka_modelas an optional input because some FASTQ may not have model information in the headers (for example downloaded from SRA).Workflow Changes ✅
Upstream Effects
None
Input Changes
medaka_modelis optional and should generally be blank to allow auto detection (avoids using the wrong model).Output Changes
The assembly version capture file has a new row recording
medaka_modelversion.Downstream Effects
Testing 🛠️
test_cov_2205_grid (new basecaller version-
r1041_e82_400bps_hac_v4.3.0) cov_2205_grid (old basecaller version -r1041_e82_400bps_hac_v4.2.0)Test(s) performed:
Ran test_cov_2205_grid with and without model auto-detection, checked outputs are the same, did not crash. Ran cov_2205_grid with model auto-detection and compared summary results to previous results with TheiaValidate.
Developer Checklist 👷♀️
Reviewer Checklist 🔍