This is a test version of Biostars. For the public version, visit https://www.biostars.org.
maSigPro edesign and error message

Hi,

I have 6 RNA-Seq samples in different time points and I would like to find DE genes changes during developmental stages. My samples are:

Sample Time Replicate Tissue Sample1 day1 1 Liver Sample2 day1 1 Liver Sample3 day7 2 Liver Sample4 day7 2 Liver Sample5 day12 3 Liver Sample6 day12 3 Liver

I am trying to understand and use “maSigPro” to find DE genes during different developmental stages.

Here is what I did:

normData <- read.delim(normCounts, row.names=1, header=TRUE);

colData= read.delim(samples, row.names=1, header=TRUE)

keep <- rowSums(cpm(normData)>1) >= 2

countdz <- normData [ keep, ]

My question is: considering above table, could you please help me to make my edesign table? I have already tried the below but I got error message when calling p.vector()…

colnames(countdz)

rownames(colData)

edesign <- as.data.frame(cbind(colData$Time, colData$Replicate))

design <- make.design.matrix(edesign, degree=2)

What does the following mean? Why there is NA in timepoint comparison?

design$groups.vector

[1] NA "V2" NA "V2"

fit <- p.vector(countdz, design, Q = 0.05, MT.adjust = "BH", counts = TRUE)

Error in dat[, as.character(rownames(dis))] : subscript out of bounds

tstep <- T.fit(fit, alfa = 0.05)

[1] "fitting gene 100 out of 363"

[1] "fitting gene 200 out of 363"

[1] "fitting gene 300 out of 363"

Error in if (group != indepen) { : missing value where TRUE/FALSE needed

Could you please help me to correct my “edesign” and also subsequent error messages?

Thanks

Simon

rna-seq
• 4,904 views • link
updated by A. Domingues A. Domingues • written by simonjean434 • simonjean434

Hi Simon, did you eventually found a solution for this problem? I am running into the same issue and would appreciate some help.

• 0 views • link
written by A. Domingues A. Domingues

1 answer

Hi had the exact same problem and eventually found the solution:

When creating the design object it needs a variable column which is not Time. My guess is that maSigPro was designed with multi-factorial experiments in mind, as seen in the vignette, and expects another (mock) condition other than time (and replicate). So in case like that of the OP, this should suffice to avoid the errors:

edesign$Tissue <- "Liver"

~Sadly this is not clear at all in the vignette.~

Edit: My bad. Section 5 of the vignette mentions this.

• 0 views • link
written by A. Domingues A. Domingues

Hello,

I have same problem. Can you please explain how to deal with this?

• 1 views • link
written by sbb • sbb

Log in to answer this question.

More posts like this