Error in R Script

Question

[786]

Hi I was performing my work on GSE71416 and got the following kind of error. Can anyone please pinpoint what does this error mean..

> rmaData = rma(gse71416preset) 

Background correcting
Normalizing
Calculating Expression
Warning message:
package ‘hgu133plus2cdf’ was built under R version 2.15.1 
> gse71416eset= exprs(rmaData) 
> groups = paste(pData(phenoData(gse71416dat[[1]]))$characteristics_ch1.1,pData(phenoData(gse71416dat[[1]]))$characteristics_ch1.2, sep=".")
> groups[groups=="Gene expression from OAT_obese and non-diabetic subjects"]="control"
> groups[groups=="Gene expression from OAT_obese and diabetic subjects"]="diabetic"
> f = factor(groups, levels=c("control","diabetic"))

Error:

> design_gse71416 = model.matrix(~0+f)
Warning message:
In is.na(rows) : is.na() applied to non-(list or vector) of type 'NULL'

It'll be very helpful if anyone can help me to remove the error..

Answer 1

[WouterDeCoster]

This is a warning, probably something is wrong. You may want to check whether your variable (f) is what you think it is.

[786]

Thannx alot I've removed it out now again there is error can u help me correcting it please it'll be very helpful fr me...

> study1 = "GSE71416" 
> gse71416dat = getGEO('GSE71416') 
Found 1 file(s)
GSE71416_series_matrix.txt.gz
trying URL 'ftp://ftp.ncbi.nlm.nih.gov/pub/geo/DATA/SeriesMatrix/GSE71416/GSE71416_series_matrix.txt.gz'
ftp data connection made, file length 3862808 bytes
opened URL
downloaded 3.7 Mb

File stored at: 
C:\Users\NADAMU~1\AppData\Local\Temp\RtmpSMMQry/GPL570.soft
> gse71416raw= getGEOSuppFiles(study1) 
[1] "ftp://ftp.ncbi.nlm.nih.gov/pub/geo/DATA/supplementary/series/GSE71416/"
trying URL 'ftp://ftp.ncbi.nlm.nih.gov/pub/geo/DATA/supplementary/series/GSE71416//GSE71416_RAW.tar'
ftp data connection made, file length 92989440 bytes
opened URL
downloaded 88.7 Mb

trying URL 'ftp://ftp.ncbi.nlm.nih.gov/pub/geo/DATA/supplementary/series/GSE71416//filelist.txt'
ftp data connection made, file length 3638 bytes
opened URL
downloaded 3638 bytes

> tarfile <- grep("\\.tar$", rownames(gse71416raw), value = TRUE)
> untar(tarfile, exdir="study1")
> celFiles <- unlist(list.files("study1", pattern = "\\.CEL.gz", full.names = TRUE)) 
> gse71416preset = ReadAffy(filenames = celFiles)
> rmaData = rma(gse71416preset) 

Background correcting
Normalizing
Calculating Expression
Warning message:
package ‘hgu133plus2cdf’ was built under R version 2.15.1 
> gse71416eset= exprs(rmaData)
> groups = pData(phenoData(gse71416dat[[1]]))$source_name_ch1
> groups=as.character(groups) 
> groups[groups=="OAT_Obese and non-diabetic"]="control"
> groups[groups=="OAT_Obese and diabetic"]="diabetic"
> f = factor(groups, levels=c("control","diabetic")) 
> design_gse71416 = model.matrix(~0+f) 
> colnames(design_gse71416) = levels(f) 
> cont.matrix = makeContrasts(diabetic-control, levels=design_gse71416)



Error:
> fit = lmFit(gse71416eset, design_gse71416)
Coefficients not estimable: control diabetic 
Error in lm.fit(design, t(M)) : 0 (non-NA) cases

[WouterDeCoster]

Could you print out the variables f and design_gse71416? To get an idea if everything is as it should be.

[786]

how can I print out variables f?Please let me know bcoz I'm naive in R.

[WouterDeCoster]

You can just type

f

and press enter. Maybe it would be good to follow an introduction/tutorial in R, that would make the rest of your analysis less painful and easier to understand (e.g. http://swirlstats.com/)

[786]

after typing f I received this what does it mean please can u explain it...

> f
 [1] <NA> <NA> <NA> <NA> <NA> <NA> <NA> <NA> <NA> <NA> <NA> <NA> <NA> <NA> <NA>
[16] <NA> <NA> <NA> <NA> <NA>
Levels: control diabetic

[WouterDeCoster]

Had to download some data to reproduce your problem, turns out the 'groups' are not what you think they are. Have a look at the groups variable after you do:

groups = pData(phenoData(gse71416dat[[1]]))$source_name_ch1

There is nothing like 'OAT_Obese and non-diabetic" or "OAT_Obese and diabetic".

I checked the variable 'pData(phenoData(gse71416dat[[1]]))' and what you need is:

groups = pData(phenoData(gse71416dat[[1]]))$title

Good luck, and I suggest to make yourself comfortable with R before proceeding.

[786]

Thank u fr your help I just changed the above command but getting the same error. Can u please help me out..

> groups = pData(phenoData(gse71416dat[[1]]))$title
> groups=as.character(groups)
> groups[groups=="OAT_Obese and diabetic"]="diabetic"
> groups[groups=="OAT_Obese and non-diabetic"]="nondiabetic"
> f = factor(groups, levels=c("nondiabetic","diabetic"))
> design_gse71416 = model.matrix(~0+f) 
> colnames(design_gse71416) = levels(f)
> cont.matrix = makeContrasts(diabetic-nondiabetic, levels=design_gse71416)

> fit = lmFit(gse71416eset, design_gse71416)
Coefficients not estimable: nondiabetic diabetic 
Error in lm.fit(design, t(M)) : 0 (non-NA) cases
> f
 [1] <NA> <NA> <NA> <NA> <NA> <NA> <NA> <NA> <NA> <NA> <NA> <NA> <NA> <NA> <NA>
[16] <NA> <NA> <NA> <NA> <NA>
Levels: nondiabetic diabetic

[WouterDeCoster]

Check your variables if something goes wrong.

The titles look like this:

OAT_Obese and diabetic [OM032007]

OAT_Obese non-diabetic [Lot OM101707]

Therefore, groups[groups=="OAT_Obese and diabetic"]="diabetic" groups[groups=="OAT_Obese and non-diabetic"]="nondiabetic" don't work because that value is not present in groups

what you need is:

groups = pData(phenoData(gse71416dat[[1]]))$title
groups=as.character(groups)
groups <- sapply(groups, substr, 1, 22)
groups[groups=="OAT_Obese and diabetic"]="diabetic"
groups[groups=="OAT_Obese non-diabetic"]="nondiabetic"
f = factor(groups, levels=c("nondiabetic","diabetic"))

In addition, you wrote "and non-diabetic" which is incorrect.