Conversion from refseq NP and SwissProt accessions to NM_ (mRNA accessions)

It turns out that automatically finding NM_ accessions for NP_ accessions is much more difficult than it should be, So I have put up a web script to do the conversion.

Try the URL: 

https://fastademo.bioch.virginia.edu/fasta_www2/NP_to_NM.cgi?acc=NP_000552,P09488
and you should get the result: 
NP_000552  NP_000552  NM_000561 
P09488     NP_000552  NM_000561
where the first field is the accession you provided, and the second and third are the refseq NP_ and refseq NM_ that match the sequence exactly. Here you can see it done on a larger scale for the a file that includes several NP_ accessions: 
for n in `cat gst_np.acc`; do 
curl https://fastademo.bioch.virginia.edu/fasta_www2/NP_to_NM.cgi?acc=$n 
done
produces: 
NP_666533  NP_666533  NM_146421 
NP_000552  NP_000552  NM_000561 
NP_001135840  NP_001135840  NM_001142368 
NP_000839  NP_000839  NM_000848 
NP_000840  NP_000840  NM_000849 
NP_000841  NP_000841  NM_000850 
NP_000842  NP_000842  NM_000851 
XP_005270842  XP_005270842  XM_005270785

The script should always work with refseq NP_'s and XP_'s. It will sometimes work with SwissProt accessions (P09488), but this is not reliable. For more reliable mapping of Uniprot accessions to RefSeq NP_'s, you need to use the Uniprot mapping service.

This script uses a database of proteins that I download from the NCBI that may not be completely up to date. So not every protein you try to map may be found. But the ones it finds should be correct.

Alternatively, I have written a Python program that uses this database at available from: 
~wrp/biol4230/proj1/NP_to_NM.py 

$ ~wrp/biol4230/proj1/NP_to_NM.py NP_000552 P09488 
NP_000552 NP_000552 NM_000561 
P09488 NP_000552 NM_000561
or 
for n in `cat gst_np.acc`; do 
~wrp/biol4230/proj1/NP_to_NM.py $n
done

It has the same "issues" as the web site, since it uses the same database.

Last modified: Friday, 30-Mar-2018 08:36:26 EDT