Puzzling Error Message While Working Through A Bioconductor Tutorial On Microarrays

Question

[Mycroft34]

I tried recently somme tutorial on microarray data analysis, using either the following link: http://bioinformatics.knowledgeblog.org/2011/06/20/analysing-microarray-data-in-bioconductor/ or the chapter on bioconductor from "R in a nutshell". After installing and loading the GEOquery package, I tried loading data as indicated:

library(GEOquery)
getGEOSuppFiles("GSE20986")

and I was returned the following error message, in both cases:

    [1] "ftp://ftp.ncbi.nlm.nih.gov/pub/geo/DATA/supplementary/series/GSE20986/"
Error in scan(file, what, nmax, sep, dec, quote, skip, nlines, na.strings,  : 
  line 1 did not have 6 elements

This is only an example, since the file involved in "R in a nutshell" is GSE2034, producing the same error. As I understand the error message, it tells me that the line 1 has a size different from the 6 elements expected for the data.frame; this is supprising for data retrieved from the NCBI server; so I think something else is faulty. Did anyone has had such an error and found what was wrong and how bypass this block. Thanks in advance.

I use R 2.15.1 on ubuntu 12.04, with bioconductor 2.10; here is the result of sessionInfo():

    R version 2.15.1 (2012-06-22)
Platform: x86_64-pc-linux-gnu (64-bit)

locale:
 [1] LC_CTYPE=fr_FR.UTF-8       LC_NUMERIC=C              
 [3] LC_TIME=fr_FR.UTF-8        LC_COLLATE=fr_FR.UTF-8    
 [5] LC_MONETARY=fr_FR.UTF-8    LC_MESSAGES=fr_FR.UTF-8   
 [7] LC_PAPER=C                 LC_NAME=C                 
 [9] LC_ADDRESS=C               LC_TELEPHONE=C            
[11] LC_MEASUREMENT=fr_FR.UTF-8 LC_IDENTIFICATION=C       

attached base packages:
[1] stats     graphics  grDevices utils     datasets  methods   base     

other attached packages:
[1] GEOquery_2.23.5     Biobase_2.16.0      BiocGenerics_0.2.0 
[4] BiocInstaller_1.4.7

loaded via a namespace (and not attached):
[1] RCurl_1.91-1 tools_2.15.1 XML_3.9-4

[brentp]

Did you check that your libcurl supports ftp? http://www.omegahat.org/RCurl/FAQ.html

[Mycroft34]

Thanks for your help; I followed what was indicated on this page:

curl -V
curl 7.22.0 (x86_64-pc-linux-gnu) libcurl/7.22.0 OpenSSL/1.0.1 zlib/1.2.3.4 libidn/1.23 librtmp/2.3
Protocols: dict file ftp ftps gopher http https imap imaps ldap pop3 pop3s rtmp rtsp smtp smtps telnet tftp 
Features: GSS-Negotiate IDN IPv6 Largefile NTLM NTLM_WB SSL libz TLS-SRP

It seems that curl support ftp; is that different from libcurl ?

[Sean Davis]

Did you get a message that looked like:

[1] "ftp://ftp.ncbi.nlm.nih.gov/pub/geo/DATA/supplementary/series/GSE20986/"
trying URL 'ftp://ftp.ncbi.nlm.nih.gov/pub/geo/DATA/supplementary/series/GSE20986//GSE20986_RAW.tar'
ftp data connection made, file length 56360960 bytes
opened URL
==================================================
downloaded 53.8 Mb

[Mycroft34]

Thanks for your help; No I didn't; I see what you mean; the double // would be the source of the problem; but I just got the message from above;

however, I remember having such a message in another circumstance; how did you solved this problem ?

For the moment, and specifically for the web tutorial (http://bioinformatics.knowledgeblog.org/2011/06/20/analysing-microarray-data-in-bioconductor/), I bypassed the block by downloading the files.

But the problem remains for the "R in a nutshell example".

Answer 1

[brentp]

This works for me with sessionInfo pasted below. Perhaps try setting LC_ALL=C to test as only our locales differ.

> sessionInfo()
R version 2.15.1 (2012-06-22)
Platform: x86_64-unknown-linux-gnu (64-bit)

locale:
[1] C

attached base packages:
[1] stats     graphics  grDevices utils     datasets  methods   base     

other attached packages:
[1] GEOquery_2.23.5     Biobase_2.16.0      BiocGenerics_0.2.0 
[4] BiocInstaller_1.4.7

loaded via a namespace (and not attached):
[1] RCurl_1.91-1 XML_3.9-4    tools_2.15.1

EDIT:

Is your libcurl built with FTP support? http://www.omegahat.org/RCurl/FAQ.html

[Mycroft34]

Thanks for your reply; I tried this LC_ALL=C settings, but the error remained.

[Neilfws]

Works for me too, with locale = en_AU.UTF-8. The error message is misleading; it just means that getGEOSuppFiles() could not access the remote file(s) for some reason. Possibly a transient network error.

[Mycroft34]

Thanks for this info; I also thought that network could be the problem; all internet traffic in our institute is passing through a proxy; could it be the cause of that error ? and how could I manage to bypass this block ?

[Sean Davis]

See the help for download.file.

[Mycroft34]

I read it and set the proxy using

export HTTP_PROXY (and  the same for FTP_PROXY)

before running R; I also checked that the proxy was set in R using Sys.getenv(), and it appeared set; but I am still having the same error:

> getGEOSuppFiles('GSE20986')
[1] "ftp://ftp.ncbi.nlm.nih.gov/pub/geo/DATA/supplementary/series/GSE20986/"
Error in scan(file, what, nmax, sep, dec, quote, skip, nlines, na.strings,  : 
line 1 did not have 6 elements

I also updated the packages, since I remembered RCurl having been updated some days ago. A new version have been installed (with several warnings, but nothing else); I am running now RCurl version 1.95-0.1.

I have also received the message :

Setting options('download.file.method.GEOquery'='curl')

after loading GEOquery. Does it means that linux curl is used instead of RCurl ?

[Raony Guimarães]

I don't know why but I'm getting the same error message when trying to use BiomaRt. I'm attaching here my error. Looks very similar with yours. Commands:

library("biomaRt")

ensembl = useMart("ensembl",dataset="hsapiensgeneensembl")

affyids=c("202763at","209310sat","207500at")

getBM(attributes=c('affyhgu133plus2', 'entrezgene'), filters = 'affyhgu133plus2', values = affyids, mart = ensembl)

Error in scan(file, what, nmax, sep, dec, quote, skip, nlines, na.strings, : line 4 did not have 2 elements

R version 2.15.1 (2012-06-22)

Platform: x86_64-pc-linux-gnu (64-bit)

locale: [1] LCCTYPE=enUS.UTF-8 LCNUMERIC=C LCTIME=enUS.UTF-8
[4] LCCOLLATE=enUS.UTF-8 LCMONETARY=enUS.UTF-8 LCMESSAGES=enUS.UTF-8
[7] LCPAPER=C LCNAME=C LCADDRESS=C
[10] LCTELEPHONE=C LCMEASUREMENT=enUS.UTF-8 LCIDENTIFICATION=C

attached base packages: [1] stats graphics grDevices utils datasets methods base

other attached packages: [1] biomaRt2.12.0 BiocInstaller1.4.7

loaded via a namespace (and not attached): [1] RCurl1.95-0.1 tools2.15.1 XML_3.95-0.1

[Mycroft34]

Since changing the RCurl version did not solved my problem (with GEOquery), I tried your case; I suggest that you edit your reply, since the name of the attribute from getBM is incorrect: it should be

"affy_hg_u133_plus_2"

instead of "affyhgu133plus2"; that makes reproducing your case a little bit difficult, since I had to retrieve the correct name.

The same apply to "hsapiensgeneensembl", that should be

"hsapiens_gene_ensembl".

Correction: to have the name correctly inserted in your message, you should put your R commands as code (inserting 4 space in front of it); otherwise, the underscores are removed (at least in the comments).

Answer 2

[Raony Guimarães]

Finally, I solved the problem!

You need to download the previous version of RCurl http://cran.r-project.org/src/contrib/Archive/RCurl/RCurl_1.91-1.tar.gz

and install using the command:

install.packages("~/Downloads/RCurl_1.91-1.tar.gz", repos=NULL)

Thank you brentp for the insights!

[Mycroft34]

Unfortunately, changing RCurl version did not correct my own problem with the GEOquery package. May be another package ought to be backed to a previous version; does anyone has an idea what package it would be ? (see sessionInfo in original message).

Answer 3

[zhanxw]

I manually download files, and then manually import the data:

gset <- getGEO(filename="GSE10246_series_matrix.txt.gz", GSEMatrix = TRUE)