broadinstitute / viral-ngs

Viral genomics analysis pipelines
Other
189 stars 67 forks source link

BLAST Database error #555

Open haowenz opened 7 years ago

haowenz commented 7 years ago

Hi,

Reported this error when using deplete_blastn_paired.

BLAST Database error: No alias or index file found for nucleotide database [xxxx.fasta] in search path

Does anyone know why?

Thanks!

tomkinsc commented 7 years ago

Have you set the config.yaml file to use a path containing a BLAST depletion database? We have a few pre-built databases described in section 4.2.3 of the docs: http://viral-ngs.readthedocs.io/en/latest/pipeuse.html#setting-up-an-analysis-directory

haowenz commented 7 years ago

Thanks, I will try.

I was faced with some other bugs when refining my assembly. Besides, some of the scripts only accept BAM files as input and then convert them back to FASTQ files before running the main part. This kind of conversion is a waste of time when dealing with large read sets.

How can I join the development of viral-ngs and help you fix them?

dpark01 commented 7 years ago

Hi @howenz , sorry a number of us have been traveling or otherwise occupied. But to your latter question, as this is a public github repo, anyone can submit pull-requests (though as noted in PR #574 there is some unfortunate round-aboutness to it.. better to PR against our "staging" branch than the "master" branch).

That being said, if you are suggesting any significant changes, it's often worth discussing them here as Issues before attempting an implementation, because some, but not all of these design patterns are often intentional. If you find that your own use cases are significantly different from ours and feel the need to diverge in design goals, you can always maintain your own fork.

The BAM file I/O for every command is highly intentional, both from a user interface standpoint as well as for functional reasons / data preservation. Time lost to conversion is sometimes significant, but is deemed to be an acceptable tradeoff.

As for your original question about the BLAST database, have you resolved that one? Make sure to run all the indexing commands on the fasta you intend to use, or use one of our pre-indexed ones.