Hi I have to identify differently expressed genes using DESeq. I have completed up to HTseq and I have HTseq output files also. I have two conditions with two replicates (arabinose Rep1,2 ,glucose Rep1,2) . When i run DESeq i got error like this ( None of your conditions is replicated. Use method='blind' to estimate across conditions, or 'pooled-CR', if you have crossed factors)
My input file is:
arabinose_1 arabinose_2 glucose_1 glucose_2
NCU00003 831 566 247 371
NCU00004 1164 964 734 1047
NCU00005 466 489 241 270
NCU00006 2527 2517 1828 1821
NCU00007 887 911 417 336
My R version 3.3.2 (2016-10-31)
This is my code: (I just copied and pasted what i get)
> library(DESeq2)
> countsTable <- read.delim ('SRP002628_readCounts.txt',header=TRUE)
> rownames(countsTable) <- countsTable$Gene_ID
> countsTable <- countsTable[,-1]
> head(countsTable)
arabinose_1 arabinose_2 glucose_1 glucose_2
1 831 566 247 371
> conditions <- factor(c("arabinose_1","arabinose_2"," glucose_1","glucose_2"))
> countDataSet <- newCountDataSet(countsTable, conditions)
> countDataSet <- estimateSizeFactors(countDataSet)
> sizeFactors(countDataSet)
arabinose_1 arabinose_2 glucose_1 glucose_2
1.1572973 1.0995991 0.8849006 0.9018904
> head(counts(countDataSet))
arabinose_1 arabinose_2 glucose_1 glucose_2
1 831 566 247 371
2 1164 964 734 1047
> head(counts(countDataSet,normalized=TRUE))
arabinose_1 arabinose_2 glucose_1 glucose_2
1 718.0523 514.7330 279.1274 411.3582
> countDataSet <-estimateDispersions(countDataSet)
**Error in .local(object, ...) :
None of your conditions is replicated. Use method='blind' to estimate across conditions, or 'pooled-CR', if you have crossed factors.**
please help me. I literally confused. I know it is a basic one. But i am not able to solve.
1 answer
From the newCountDataSet help about the condition argument:
A factor of experimental conditions (or treatments, or tissue types, or phenotypes, or the like). The length of the factor has to be equal to the number of columns of the countData matrix, assigning a condition to each sample.
And this is how it is used in the example:
cds <- newCountDataSet(countsTable, c( "A", "A", "A", "B", "B" ) )
So, I believe you want in your code:
conditions <- factor(c("arabinose", "arabinose", "glucose", "glucose"))
i have to calculate RPKM values for DESeq
No, you definitely don't have to do that. Input to DESeq2 should be raw read counts.
Log in to answer this question.
Please edit and provide a more informative title. Copying and pasting the error message in the title provides no hints/useful information.