This is a test version of Biostars. For the public version, visit https://www.biostars.org.
Error in R Script

Hi I was retrieving differentially expressed genes through R and got an error at the very last step. Please help me out in removing following error it will be very helpful.

Error
results= topTable(fit2, adjust ='BH', number = nrow(gse32269eset))
Error in (function (..., row.names = NULL, check.rows = FALSE, check.names = TRUE, :
arguments imply differing number of rows: 54675, 22283
r affy limma bioconductor
• 4,839 views • link
updated by ATRX ATRX • written by 786 786

I think It's related to selecting the number of DEGs,why you select all the prob-id in your GSE? try with a number,lets say 100, I suggest you specify log fold change for your topranked gene e.g logFC=0.05

• 0 views • link
written by Shamim Sarhadi Shamim Sarhadi

1 answer

try using: topTable(fit2, adjust ='BH', number = Inf)

I think you are filtering some probes or aggregating the probes due to which there is a difference in the number of rows between fit2 and nrow(gse32269eset)

• 0 views • link
written by ATRX ATRX

I got the same error by running this command..Can you please help me more?

• 1 views • link
written by 786 786

I am not sure but it is working for me. I presume that you are using GSE32269

> exprs = exprs(eset) 
> dim(exprs)
[1] 22283    55
> design <- model.matrix(~ type + Tmprss2.erg + 0, as.data.frame(group))
> head(design)
  typePrimary typeMet typeNormal Tmprss2.ergPositive
1           1       0          0                   0
2           1       0          0                   0
3           1       0          0                   0
4           1       0          0                   0
5           1       0          0                   0
6           1       0          0                   0
> cont.matrix <- makeContrasts(NormalvsPrimary = typePrimary-typeNormal, levels=design)
> print(cont.matrix)
                     Contrasts
Levels                NormalvsPrimary
  typePrimary                       1
  typeMet                           0
  typeNormal                       -1
  Tmprss2.ergPositive               0
> fit1 <- lmFit(exprs, design)
> fit2 <- contrasts.fit(fit1, cont.matrix)
> fit2 <- eBayes(fit2)
> results <- topTable(fit2, "NormalvsPrimary", number=Inf)
> head(results)
                logFC   AveExpr         t      P.Value    adj.P.Val        B
201596_x_at  6.359687 10.604378  18.36430 9.175354e-25 2.044544e-20 45.75283
215054_at   -1.735632  5.432611 -17.25748 1.593483e-23 1.775379e-19 43.02282
211820_x_at -2.799107  4.629426 -16.60052 9.224250e-23 6.374164e-19 41.33593
206207_at   -4.621926  6.161970 -16.52111 1.144220e-22 6.374164e-19 41.12856
206676_at   -5.027171  4.179033 -16.01332 4.615017e-22 2.056728e-18 39.78449
204959_at   -5.102275  4.839911 -15.76017 9.351825e-22 3.473112e-18 39.10262
• 1 views • link
written by ATRX ATRX

It worked but I need gene names too in my results how can I get them??

• 1 views • link
written by 786 786

You need annotation file from Affymetrix and then map it to the probes. You may also use David to convert them to gene ids/gene symbol.

• 1 views • link
written by ATRX ATRX

Thank u so much for our help.

• 1 views • link
written by 786 786

Thanx alot fr your help.

• 1 views • link
written by 786 786

Sorry I have ovary cancer samples stage 1, 2, 3 and normal samples I was going to have something like your above code resulted in P.value , adj.P.Val , logFC

I did like below

library(limma)

library(affy)

Data<-ReadAffy()

eset<-rma(Data)

exprs<-exprs(eset)

dim(exprs)

[1] 54675 4

design <- model.matrix(~ type + Tmprss2.erg + 0, as.data.frame(group))

Error in as.data.frame(group) : object 'group' not found

Actually I was going to compare normal with rest of cancers in 3 stages but I got error

how I can have your top table please?

• 1 views • link
written by zizigolu zizigolu

Com'on.. you need to define the group first.. i.e. which sample belongs to which group. Read about limma first and try to understand the code. Please think and apply. I am sure you don't want yourself to be spoon-fed.

• 1 views • link
written by ATRX ATRX

It is never a good idea to ask a new unrelated question in an existing thread. Please start a new thread.

• 1 views • link
written by GenoMax GenoMax

Sorry If I ask such a question in a new thread the Admins will close my post abruptly because they supposed me as a spoon-feed OP :)

• 1 views • link
written by zizigolu zizigolu

Log in to answer this question.

More posts like this