Search
Question: Obsolete REFSEQ annotation
0
8 months ago by
Ed Siefker220
United States
Ed Siefker220 wrote:

I have reads counted with rna.fa.gz from:
ftp://ftp.ncbi.nih.gov/genomes/M_musculus/ARCHIVE/BUILD.37.1/RNA/

It's labeled with IDs that look like this:

gi|142348699|ref|NM_010886.2|

I've extracted the REFSEQs, and used org.Mm.eg.db to get the SYMBOLs.

X1      X2
1 gi|126352347|ref|NM_028260.2|  Immp1l
2 gi|142348699|ref|NM_010886.2|  Ndufa4

Unfortunately, a number of REFSEQs seem to be obsolete.

> select(org.Mm.eg.db, keys="XM_001478046", keytype="REFSEQ", columns="SYMBOL")
Error in .testForValidKeys(x, keys, keytype, fks) :
None of the keys entered are valid keys for 'REFSEQ'. Please use the keys method to see a listing of valid arguments.

https://www.ncbi.nlm.nih.gov/nuccore/XM_001478046

NCBI Reference Sequence: XM_001478046.1 (click to see this obsolete version)

How do I use Bioconductor to convert XM_001478046 to NM_001142441(and then Sap18b)?

written 8 months ago by Ed Siefker220

I could also try to annotate using the GI number.  Are there annotation packages that use the GI number?  It doesn't seem to be a column in org.Mm.eg.db.

> columns(org.Mm.eg.db)
[1] "ACCNUM"       "ALIAS"        "ENSEMBL"      "ENSEMBLPROT"  "ENSEMBLTRANS"
[6] "ENTREZID"     "ENZYME"       "EVIDENCE"     "EVIDENCEALL"  "GENENAME"
[11] "GO"           "GOALL"        "IPI"          "MGI"          "ONTOLOGY"
[16] "ONTOLOGYALL"  "PATH"         "PFAM"         "PMID"         "PROSITE"
[21] "REFSEQ"       "SYMBOL"       "UNIGENE"      "UNIPROT"

Could you add the output of sessionInfo() to your question?

This is something that has been asked for in the past. We provide the most current ID mappings from NCBI, but all of the revision history is opaque to the annotation packages.

I don't think it would be insurmountable to add the Gene ID history (it's provided at ftp.ncbi.nlm.nih.gov/gene/DATA/gene_history.gz) The revision history for RefSeq might be parseable from here (ftp://ftp.ncbi.nlm.nih.gov/refseq/removed/), although that's lots of data.

Anyway, there is some interest from people who have old data, but who want to annotate using their current Bioconductor installation. The alternative of trying to install successively older versions of R/BioC in order to match with the era of your underlying data is probably a suboptimal strategy.

The revision history might also be accessed using RCurl and querying nuccore, with a correctly formatted URI, like https://www.ncbi.nlm.nih.gov/nuccore/XM_001478046?report=girevhist, which shows the history for the RefSeq ID that Ed cares about.