Heebo-->Gene Symbol Conversion, Or A Better Way To Get Geo Annotation Files

Question

[ellenmseb]

Hello, I'm trying to do a meta-analysis of separate microarray studies conducted on different platforms. So I need all platforms' probes to map to common gene IDs. Problem is, for some of the studies, none of the GEO annotation files contain any identifier that identifier-conversion tools recognize. None of the tools on Gene Id Conversion Tool have been able to recognize or convert any of these columns.

This is the format of GEO's SOFT files, with no commonly recognized identifiers:

> hCP000001    NM_001101.2    GATTGCACATTGTTGTTTTTTTAATAGTCATTCCAAATATGAGATGCATTGTTACAGGAAGTCCCTTGCC    chr7_-_5340026_5343473    gi|5016088|ref|NM_001101.2||gi|5016088|ref|NM_001101.2||+|1193|1793_228    actin,
> beta    
> 
> hCP000286    NM_012423.2    CGGACTCCTGGTCTGAGCCCAATAAAGACTGTTAATTCCTCATGCGTTGCCTGCCCTTCCTCCATTGTTG     chr19_+_54682675_54687374    gi|14591905|ref|NM_012423.2||gi|14591905|ref|NM_012423.2||+|542|1142_77    ribosomal
> protein L13a

Does anyone know of a conversion tool for any of the above identifiers, or how to get gene-symbol probe annotations from GEO?

Thanks very much in advance!

[Neilfws]

There are at least 2 "commonly recognized identifiers" there; RefSeq mRNA and Entrez UID (GI). You can easily map the former to other gene IDs using e.g. BioMart.

Answer 1

[Sudeep]

I see that you have some refseq ids in your probe mappings (the ids starting with NM_.*). Assuming that these studies are fairly recent may be you can map these refseq ids to Entrez gene ids. You can parse refseq id to Entrez gene id mappings from this file