search

emepyc / Blast2lca

Calculates the lowest common ancestors of each query sequence in a Blast result
GNU General Public License v2.0
31 stars 9 forks source link

README #2

Closed nick-youngblut closed 10 years ago

nick-youngblut commented 10 years ago

Some confusion I had with the README file:

1) In the README: -levels superkingdom:phylum:class:family In blast2lca help: -levels='superkingdom:phylum:class:family' The 2nd version seems to work, while the 1st doesn't and furthermore it doesn't throw any errors.

2) The README setup is confusing for trying to use blast2lca for a nucleotide query. You only cover 'gi_taxid_prot.dmp.gz' in the uncompression and gitaxid2bin steps, but not ''gi_taxid_nucl.dmp.gz' Using the wrong *.bin file won't throw any errors, but all taxonomic labels will be 'unknown'. This combined with the warnings when using 'blastn -outfmt 6' was pretty confusing at first.

emepyc commented 10 years ago
  1. Go's flag package is a bit restricted with the format of the arguments it understands. You need to use the equal as stated in the tool's help. I have fixed the README
  2. I thought it was obvious that if you have run your blast against a nucleotide database you should file that provides mappings between nt GIs and taxonomy IDs. Anyway, I have added an explicit node on that in the README.
nick-youngblut commented 10 years ago

The prot|nuc is pretty obvious. I was multitasking trying to set up blast2lca while to some other work, which caused me to make the error at first. By making the README more explicit, maybe others with also not make the same stupid mistake.

Thanks for updating the README. I really like the features of blast2lca, especially it¹s speed.

Nick

From: Miguel Pignatelli notifications@github.com Reply-To: emepyc/Blast2lca <reply+i-28741445-a8a274ffd1ffa25393da9d51bf6a68c29a182441-2468572@reply.git hub.com> Date: Wednesday, March 5, 2014 at 6:21 AM To: emepyc/Blast2lca Blast2lca@noreply.github.com Cc: Nicholas Youngblut nyoungb2@gmail.com Subject: Re: [Blast2lca] README (#2)

  1. Go's flag package is a bit restricted with the format of the arguments it understands. You need to use the equal as stated in the tool's help. I have fixed the README
  3. I thought it was obvious that if you have run your blast against a nucleotide database you should file that provides mappings between nt GIs and taxonomy IDs. Anyway, I have added an explicit node on that in the README.

‹ Reply to this email directly or view it on GitHub https://github.com/emepyc/Blast2lca/issues/2#issuecomment-36732749 .