Difficult To Download Gene Sequences From Ncbi

Question

[jcastrofigueroa]

Hello everyone: I'm having a problem trying to download gene sequences from the Gene database at NCBI website using biopyhon. I iniciated the code by setting up a basic test search for two gene sequences in the "gene" database for S. coelicolor (txid100226).

from Bio import Entrez
Entrez.email = "[e-mail]"
handle = Entrez.esearch(db="gene",term="txid100226[Organism]",retmax=2)
record = Entrez.read(handle)

The first ID for the first hit on this search is:

record_list = record["IdList"]
print record_list[0]
1096915

So this first ID was used to download the gene of interest by using this:

seq = Entrez.efetch(db="gene",id=record_list[0],rettype="fasta").read()

However the result stored in "seq" is the following:

http://www.ncbi.nlm.nih.gov/data_specs/dtd/NCBI_Entrezgene.dtd">
<Entrezgene-Set>
SCO1489 –DNA-binding protein [Streptomyces coelicolor A3(2)] 
DNA-binding protein
    
Other Aliases: 
SCO1489, SC9C5.13, bldD
Genomic context: 
Chromosome
 Annotation: 
NC_003888.3 (1592381..1592884)
ID:
 
1096915
 
 
</Entrezgene-Set>

If I put db="protein" instead of gene I get the correct protein sequence.

I realize that one way to download the DNA sequence was manually, directly from the contig NC_003888.3 in S. coelicolor at the position 1592381..1592884 for this particular ID. That info is stored in "seq"

So here is the question: Is there any method (or trick) to download that DNA sequence using biopython? How can I solve this problem?

JFC

Answer 1

[Neilfws]

The short answer is that rettype = "fasta" is not a valid return mode for the Gene database. Please refer to Table 1 in the EFetch section of the NCBI EUtils documentation.

The longer answer - how to solve this problem - I'll edit this answer later, no time to write it just now.

[jcastrofigueroa]

Even if I try to change the rettype, it doesn't work. The gene sequence for this example is within contig sequence, so the GI code for this sequence directs you to the contig. I don't know what to do to solve it, but thank you for your answer.

[Neilfws]

Well no, changing rettype won't work. The only valid rettype for db=Gene is gene_table; valid retmodes are asn.1, xml and text. In short: sequences cannot be retrieved from the Gene database.

Answer 2

[Ashutosh Pandey]

Well I am not used to using Entrez gene but I think you are retrieving the Entrez gene page information instead of the sequence information. You should try either "genbank" or "nucleotide" instead of "gene" and see if it helps.

[jcastrofigueroa]

Thanks for your answer, but it didn't work :( If I use "gene bank" it displays an error and if I try with nucleotide database, what I get is the whole contig. Hmm, about using Entrez gene I'm sure that I'm not retrieving the information page, because I get a protein sequence.

Answer 3

[Leandro Lima]

Hello! I think this could help you.

problem when downloading large number of sequences from Genbank

[Neilfws]

Not really since fasta cannot be retrieved from the Gene database.

[Leandro Lima]

In this case, db="nuccore"