Question: DESeq2 Not Recognizing that my data set has been transformed when inputting. Help!
gravatar for alyssamariemarron
5 weeks ago by
alyssamariemarron0 wrote:

Hi I am trying to do a DESeq2 analysis. My PI previously batch corrected and vst transformed our gene expression data prior to sending it to me. I am trying to load it into a deseq data set but the code I am using it recognizing it as a count data set not a transformed data set. (i.e. when I try to run DESeq on my data set, it tells me it is not in integers, which is because it has been vst transformed). I believe the error is because the DESeqDataSetfromMatrix function input the gene expression data as count data. My question is what input method do I use instead to load the transformed matrix into a DESeq data set that keep it recognized as a transformed object?

Here is my full code up to the point of trying to create the DESeqDataSet

#Loading Geek Gene Expression data
load("GEEK-ID 11576 genes outliers removed.RData")
#Moving the file from a data frame to a matrix, we are also using the t() function to transform the matrix so it has participant numbers along the top and gene names along the side
DESdata <- as.matrix(t(GEEKexpressionData))
# Now I will load the metadata (moved the file into the pasilla package extdata to make it so I didnt need to alter the code)
pasAnno <- system.file("extdata",
                       "Geek Data for DESeq2.csv",
                       package="pasilla", mustWork=TRUE)
coldata <- read.csv(pasAnno, row.names=1)
#Now we will use the match function to make sure the rows of coldata match the columns of desdata
#This created a vector that contained all the participant numbers in the coldate that had removed all races except caucasian and african american, and also removed CRP outliers
#This now created a new matrix of the gene expression data, the participant columns have been made to match the rows of the coldata
#There are NA columns present in the new matrix newDESdata that represent the that were previously outliers removed from the gene expression data 
#Next we will create a new matrix without these NA values
newDESdata2 <- newDESdata[,
ADD COMMENTlink modified 5 weeks ago by genomax87k • written 5 weeks ago by alyssamariemarron0
gravatar for swbarnes2
5 weeks ago by
United States
swbarnes28.2k wrote:

I don't think you can import a vst transformed dataset like that. For one thing, when you run vst, you make a whole new DESeq Transform object out of it. It's not a slot that is added to the original dds object. I don't think that you can untransfrom vst counts, which is what you would need to find DE genes.

The transformed data can be used for visualization, but I don't think you can do much else with it. You need to get the raw data.

You usually don't really batch correct data at the count level anyway. The preferred way is to add the batch as an element in the design.

ADD COMMENTlink written 5 weeks ago by swbarnes28.2k
Please log in to add an answer.


Use of this site constitutes acceptance of our User Agreement and Privacy Policy.
Powered by Biostar version 2.3.0
Traffic: 737 users visited in the last hour