[BioC] R: Help request about GEOquery and GSE9820

[Manca Marco (PATH)]

Dear Sean,

thank you for your kind and prompt answer.

I will check immediately my typo (the gse instead of gse9820)! =P

I have used this command
> gse9820<-getGEO("GSE9820", GSEMatrix=FALSE, destdir="C:\\Documents and Settings\\Marco Manca\\Mijn documenten\\GSE9820")

instead of
gse9820 <- getGEO('GSE9820')

because for some reason it wasn't importing/parsing the individual records and asking for
>gse9820
was giving a weird answer full of N/A signatures... However I will give it a shot later on and will give you a feedback on it.

Thank you again and keep up the great work!

Best regards,
Marco

________________________________________
Da: seandavi at gmail.com [seandavi at gmail.com] per conto di Sean Davis [sdavis2 at mail.nih.gov]
Inviato: giovedì 16 luglio 2009 13.36
A: Manca Marco (PATH)
Cc: Bioconductor Newsgroup
Oggetto: Re: Help request about GEOquery and GSE9820

On Thu, Jul 16, 2009 at 5:07 AM, Manca Marco (PATH) <m.manca at path.unimaas.nl<mailto:m.manca at path.unimaas.nl>> wrote:


Dear Sean,

good morning. I hope my email finds you well.

I apologize for incomodating you with a puny question about the use of your nice GEOquery packace, but I am quite a new user of it (Alberto Goldoni, who is a good friend of mine and a quite active member of Bioductor community whom you may know maybe) and I can not find a solution to an error message which is blocking me since one whole day. I hope in your understanding and support.

The case is as follows:
- I am trying to access the GSE9820 record on Gene Expression Omnibus.

- Using the command
> gse9820<-getGEO("GSE9820", GSEMatrix=FALSE, destdir="C:\\Documents and Settings\\Marco Manca\\Mijn documenten\\GSE9820")

I succeeded to acquire the data (I will omitt the result of the query gse9820 since it's some tens of thousands line which would just impair our commnication here)
- The command
> Table(GSMList(gse9820)[[1]])[1:5, ]
gives me the following output (which looks nice I think)
     ID_REF       VALUE
1 ILMN_10000  6.88027374
2 ILMN_10001 12.36707372
3 ILMN_10002  2.41509385
4 ILMN_10004 6.898955513
5 ILMN_10005 3.728241841

- But the command
> Columns(GSMList(gse9820)[[1]])[1:5, ]
gives me this (which looks nasty)
    Column                      Description
1    ID_REF
2     VALUE log2 normalized signal intensity
NA     <NA>                             <NA>
NA.1   <NA>                             <NA>
NA.2   <NA>                             <NA>

This looks this way because you asked for 5 column descriptions (1:5), but there are actually only two columns.  Doing:

Columns(GSMList(gse9820)[[1]])

will give a prettier answer, but the data is fine, either way.


- and in fact I can't manage to transform the data in an Expression set, the command
data.matrix <- log2(do.call("cbind", lapply(GSMList(gse), function(x) {
tab <- Table(x)
mymatch <- match(probesets, tab$ID_REF)
return(tab$VALUE[mymatch])
})))
giving me the following error message
Error: object 'gse' not found
Error in GSMList(gse) :
 error in evaluating the argument 'object' in selecting a method for function 'GSMList'

This is a simple cut-and-paste problem.  You called your GSE "gse9820" and not "gse".


I am sure this is due to my short experience with your package but I can not find a solution (neither in my mind nor in the previous thread in the mailing list) alone and I am desperately stuck at this point. Could you give me any hint? Even just a suggestion of something to read to become able to find a solution my self would be a much apreciated help, if not a straight answer.

The more straightforward way of doing all this is to simply use the GEOquery default (at least for any newer version of GEOquery):

gse9820 <- getGEO('GSE9820')

This will return a __LIST__ of ExpressionSets and parses only the GSE Series Matrix file (so it is much faster than using GSEMatrix=FALSE).   Typically, there is only one member in that list (and that is the case for GSE9820, so to get a single ExpressionSet, do the following:

gse9820eSet <- gse9820[[1]]

Hope that helps.

Sean


Thank you in advance for your attention and support. My best regards,
Marco



--

Marco Manca, MD

University of Maastricht
Faculty of Health, Medicine and Life Science (FHML)
Cardiovascular Research Institute Maastricht (CARIM)
E-mail: m.manca at path.unimaas.nl<mailto:m.manca at path.unimaas.nl>

"Alma Mater Studiorum" University of Bologna
<<GC Descovich>> Atherosclerosis and Metabolic Diseases Study Centre
Dept. of Clinical Medicine and Applied Biotechnology
E-mail: m.manca at unibo.it<mailto:m.manca at unibo.it>
Website: http://spazioinwind.libero.it/marco_manca_1979

General Secretary of the International Society for Mathematical Models in Medicine
E-mail: medmath0 at yahoo.it<mailto:medmath0 at yahoo.it>

General Secretary of the Grupo Luso-Italiano de Arteriosclerose
E-mail: glia.secretary at gmail.com<mailto:glia.secretary at gmail.com>
Website: http://digilander.libero.it/Grupo_Luso_Italiano/