How To Associate Cpg Coordinate With The Gene Names Or Entrez Id

Question

[hdy]

I am dealing with methylation data and see those entries are describe as CpG/CpH coordinate (UCSC hg19, Feb 2009), like: cg00013618 and cg00027083. How can I retrieve the gene names or ids from these coordinate information?

[farzaneafzali]

The Illumina's website does not work for me and the files would not be downloaded, so people can download the manifest files from supplementary sections in corresponding GPL's page. For example for GPL13534, you can download the files via this link: https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GPL13534

[GenoMax]

If you are referring to @Kevin's link it is still working fine.

[farzaneafzali]

The downloading process does not start for me so I have trouble with that link.

[Kevin Blighe]

Where are you based? Your country may be sanctioned by the US Government. NCBI is hosted at NIH, a US government agency. The other website, Illumina.com is then obviously a US-based website, too.

[farzaneafzali]

Yeah that's true. I am from Iran :/ I am truly sorry for the science that is imprisoned by the politics. But, scientists should help each other and pass the illogical borders, hence I suggested to the others like me that can download the files via that path as well. By the way, you can retrieve the genes symbol of these GPLs from the microarray data in R through sub setting the "symbol" in final steps of analyzing. The code would be something like this:

tT <- subset(tT, select=c("adj.P.Val", "P.Value", "logFC","Symbol","B", "t"))

[haskankaya]

Thank you for this! The Illumina website was not working for me (download link said that I did not have access) but the supplementary file available at this GEO site allowed me to download and use it as a lookup table (cg# --> genomic coords).

Answer 1

[Kevin Blighe]

Same issue here and just became really [really] tired of these annotation packages becoming outdated.

You can easily obtain the 450k methylation annotation data from Illumina's website, where it should always remain and not vanish into thin air:

• https://support.illumina.com/array/array_kits/infinium_humanmethylation450_beadchip_kit/downloads.html

Kevin

Answer 2

[GANI]

I see that based on the probe Identifiers, you were using Illumina Methylation data. What specific platform are you using (27k or 450k)? Also, what tools/language are you using to perform the analysis?

If you are working in R, you can use the following package to retrieve the corresponding gene information for 27k platform.

http://www.bioconductor.org/packages/release/data/annotation/html/IlluminaHumanMethylation27k.db.html

or the following package for 450k platform

http://www.bioconductor.org/packages/2.13/data/annotation/html/IlluminaHumanMethylation450k.db.html

[sjorsvanheuveln]

Here's the R-code.

source("https://bioconductor.org/biocLite.R")
biocLite("IlluminaHumanMethylation450k.db")

library(IlluminaHumanMethylation450k.db)
CpG_annotation <- as.list(IlluminaHumanMethylation450kSYMBOL[mappedkeys(IlluminaHumanMethylation450kSYMBOL)])

[hdy]

I am using 27K now but probably will also head into 450K as well, thank you for your advice and I do use R. Do you know, if I am dealing with 450k, what tools may help and give me results?

[GANI]

I just updated my post with the 450k annotation as well.

Answer 3

[Isaac Joseph]

So, for posterity, to save time and do something roughly equivalent, just do biocLite('FDb.InfiniumMethylation.hg19')instead of the above.

Details: the above appears to be deprecated; I received the following error:

    >biocLite("IlluminaHumanMethylation450k.db")

...

    Error : .onLoad failed in loadNamespace() for 'IlluminaHumanMethylation450k.db', details:
      call: NULL
      error: 'fun' is defunct.
    Use 'FDb.InfiniumMethylation.hg19' instead.
    Use 'FDb.InfiniumMethylation.hg18' instead.
    Use 'mapToGenome() function in minfi or methylumi' instead.

, suggesting that I use one of these other databases.

[Isaac Joseph]

Plot twist: said posterity now includes me, as I forgot how to do this, and moreover, forgot that I knew how to do this, so Google®ed it and came upon this question again.

[demoraesdiogo2017]

did you solve it? I managed to install FDb.InfiniumMethylation.hg19 after some suffering but I can not make something equivalent to

CpG_annotation <- as.list(IlluminaHumanMethylation450kSYMBOL[mappedkeys(IlluminaHumanMethylation450kSYMBOL)])

to work