How to fetch any older version ACCs from NCBI using programming language?

[Brace]

For example: I want to get the genomic coordinates of NR_046018.1, which has been updated and removed from the current release of NCBI databases.
Then，how can I retrieve the record using EUtils like API?

Or maybe, anyone can supply another fresh way to deal with this kind of work?
Any suggestions?

Thanks in advance.

[Brace]

Saddly, no answer, then maybe I should do it another way, does anyone know where can I find an accession number dependent version update notes on NCBI?

e.g. NM_001604.4 -> NM_001604.5 ... anything changed?

[cliffbeall]

In BioPython you can specify the accession with the decimal points and get the two different records:

>>> from Bio import Entrez
>>> Entrez.email = '[email protected]'
>>> handle1 = Entrez.efetch(db="nucleotide", id = "NR_046018.1", rettype = "gb")
>>> print handle1.read()
...older version - sequence ends gtttct...
>>> handle2 = Entrez.efetch(db="nucleotide", id = "NR_046018.2", rettype = "gb")
>>> print handle2.read()
...newer version - sequence ends gtttctg...

That is the only eutils wrapper I have experience with but I would think anything would work like that.

[Brace]

Thanks cliffbeall
I understand that EUtils can fetch older version of ACC from nucl db of NCBI, even in
genbank or xml format, but I cannot find the genomic coords from any of the returned files.
The genomic coords of current version of ACC can be obtain by esearch & efetch from db:gene element:Gene-commentary_genomic-coords easily, But this's not available for older version.

I know that the final solution is to run `splign` from sequences, but it's rather a tough way for whole genome and thousand of genes.

Any other suggestions?

[cliffbeall]

Sorry, I guess I misunderstood the question.

I don't know how to do what you want, it has to do with links between nucl and gene databases.

[Brace]

It's not your fault, my misleading indeed.
Thanks cliffbeall.
Waiting for any further suggestions ...