We were seeing sample failure events due to VADR throwing an error for long sample names: ERROR, at least one sequence name exceeds the maximum GenBank allowed length of 50
VADR should not cause the pipeline to fail. Rather, the errors should be caught and propagated in .json while the pipeline continues. This PR addresses these issues as follows:
Updating the vadr_options to use the --noseqnamemax flag so that it will not fail for long sample names
Updating the Vadr step to include better error handling that will correctly capture the error more accurately and allow the pipeline to continue
Added associated unit tests to 1. use the long sample names as a proxy for vadr failures to test the error handling, 2. ensure that the flag successfully eliminates this particular error.
This was tested via unit testing as well as through application to two separate samples.
We were seeing sample failure events due to VADR throwing an error for long sample names:
ERROR, at least one sequence name exceeds the maximum GenBank allowed length of 50
VADR should not cause the pipeline to fail. Rather, the errors should be caught and propagated in .json while the pipeline continues. This PR addresses these issues as follows:
vadr_options
to use the--noseqnamemax
flag so that it will not fail for long sample namesThis was tested via unit testing as well as through application to two separate samples.