Validation when RefSeq data are not available for CLARK

Scenario:

1. Make sure:
   • taxonomy data are available
   • "viral_database" or "bacterial_viral_database" is available
2. RefSeq data are NOT available
3. Open a valid workflow with CLARK-classify.
4. Run the workflow with the default parameters.
   Expected state: the workflow ran successfully.
5. Change some CLARK algorithmic parameters, e.g. "Minimum k-mer frequency", "Gap", etc.
6. Run the workflow again.
   Current state: CLARK attempts to build a database with new parameters. As RefSeq data are not available, the database does not contain any meaningful data. There are a bunch of errors in the log like:

   Failed to open ../../refseq/viral/AC_000001.1.fna
   Failed to open ../../refseq/viral/AC_000002.1.fna
   Failed to open ../../refseq/viral/AC_000003.1.fna
   ...
   

   But this is not reflected somehow on the WD dashboard.
   Expected state: the workflow should be considered as invalid in this case. An error should be shown on the workflow validation:

   Reference database for these CLARK settings is not available. RefSeq data are required to build it.
   

7. Put the RefSeq data to the appropriate location.
8. Run the workflow again.
   Expected state: the new database is build. There are no errors in the log.