How To Understand The Description Of Illumina Human610-Quad V1.0 Beadchip On Geo?

Question

[tflutre]

I am interested in a genotype dataset on GEO that was obtained using the Illumina Human610-Quad v1.0 BeadChip (GPL8887). Thanks to the R package GEOquery, I can easily retrieve the data in R:

library(GEOquery)
gpl <- getGEO('GPL8887', destdir='.')
head(Table(gpl), n=2)
      ID Chr   MapInfo SNP_ID
1 200003   9 139026180 200003
2 200006   9 139046223 200006

But what does this mean?

• Does the column SNP_ID correspond to the rs identifier in dbSNP? Apparently no, as rs200003 is not on chr9.
• Does the column MapInfo correspond to the SNP coordinate? If yes, on which version of the human genome reference sequence is it mapped? hg18 or hg19?

I search on the Internet (on Illumina's web site, etc) but couldn't find any information. Any idea?

[tflutre]

For info, this platform has 598,821 markers with a "rs" identifier: length(grep("rs", Table(gpl.geno)$SNP_ID))

Answer 1

[Neilfws]

1. Some of the values in SNP_ID do correspond to dbSNP, others do not. If you examine some of the later rows in the table, you'll see all kinds of identifiers used for probes with prefixes such as: rs, cnvi, and GA.
2. Yes, MapInfo is the chromosomal coordinate and since this is an older platform (2009), hg18 is the genome version.

You might find it useful to download the PennCNV package and look at the file lib/hhall.hg18.pfb - it contains probe IDs, chromosomes and coordinates which should match most of those from your platform.

[tflutre]

@neilfws Thanks, I accepted your answer. But I wanted to mention that for my purpose the file "hhall.hg18.pfb" in PennCNV only contains information redundant with the data from GEO.

[Neilfws]

True. I only mention the file since it's derived from the Illumina platform metadata, to illustrate that the probe IDs are in fact "official" Illumina IDs.

Answer 2

[Alexbmp]

I've looked up many sites of GPL8887 and it seems to be hg18 coordinates.

However, what I'm wondering now is, how can one be sure that GPL8887 is hg18; since I cannot find the documentation.

Does anybody know where to find the documentation about GPL8887 genome version information?