[BioPython] FASTA parsing errors

[Aaron Zschau]

That was the answer I was looking for, I figured something small had 
changed with the new version and the defaulting to FASTA no longer 
being the case was what was throwing things off. I have to sort out 
some new errors now but this definitely got me going along on my way.

thanks much for the detailed reply.

Aaron Zschau

On Aug 3, 2004, at 6:26 PM, Brad Chapman wrote:

> Hi Aaron;
>
> Aaron:
>>> This is the file that is being read. I know it worked in 1.24 just 
>>> fine
>>> but maybe something changed in the versions that make it not like 
>>> this
>>> format
>>>
>>> LOCUS       XM_414447               2107 bp    mRNA    linear   VRT
> [....]
>
> Jon:
>> I don't think that file conforms to the fasta format:
>> see http://ngfnblast.gbf.de/docs/fasta.html
>> I could be wrong though.
>
> Right. That's a GenBank file, which is why the Fasta parser is
> choking on it (the error message should be a lot nicer, for sure).
>
> You have two options:
>
> 1. Use a GenBank parser.
>
> 2. Retrieve Fasta sequences. Going from the code you posted
> previously, you could retrieve your search in FASTA format with the
> following:
>
> from Bio import GenBank
> from Bio import Fasta
>
> ncbi_dict = GenBank.NCBIDictionary("nucleotide", "fasta",
>              Fasta.RecordParser())
>
> seqrecord = ncbi_dict["6273291"]
>
> genbank_file = open(data_path_prefix + file_unique_id + 'fasta',
> 'w')
> genbank_file.write(seqrecord + "\n")
> genbank_file.close()
>
> This may have changed with the most recent release because the
> default for GenBank retrieval used to be fasta. Because of changes
> at NCBI this had to be updated, and I believe now defaults to
> GenBank. So, if you didn't specify "fasta" as the second argument,
> that's probably now why you are getting GenBank data. Hopefully this
> small change in your code will fix everything.
>
> Hope this helps.
> Brad
> _______________________________________________
> BioPython mailing list  -  BioPython at biopython.org
> http://biopython.org/mailman/listinfo/biopython