[BioC] multiple level factors contrast in nbinomLRT, DESeq2.

sh. chunxuan hibergo at outlook.com
Wed Aug 20 12:37:58 CEST 2014

Dear list, 
I am not sure the correct way to interpret the nbinomLRT results in multiple level factors condition.
Here is a toy example. I would to find DE genes after controlling batch effect in the experiments, in which there are multiple levels. 
## make data library(DESeq2)dds <- makeExampleDESeqDataSet(m = 18)colData <- data.frame(row.names = rownames(colData(dds)), sample = colData(dds)$sample, condition = rep(LETTERS[1:6], each = 3), batch = factor(rep(c(1, 1, 2), 6)))dds <- DESeqDataSetFromMatrix(counts(dds), colData = colData, design = ~ batch + condition)dds <- estimateSizeFactors(dds)dds <- estimateDispersions(dds)## LRTestdds <- nbinomLRT(dds, reduced = formula(~ batch ))resultsNames(dds)[1] "Intercept"        "batch_2_vs_1"     "condition_B_vs_A" "condition_C_vs_A" "condition_D_vs_A" "condition_E_vs_A" "condition_F_vs_A"res.1 <- results(dds, name = "condition_B_vs_A")res.2 <- results(dds, contrast = list("condition_B_vs_A", "batch_2_vs_1")
Questions:1.) I would like to get the DE genes between B and A, while controlling for the batch effect. Should I take "res.1" or "res.2", or both are wrong? and what is the correct way to do it?
2.) Why "res.3 <- results(dds, contrast = c("condition", "A", "b"))" gave error: "Error in normalizeSingleBracketSubscript(j, x) :   subscript contains invalid names"
3.) In order to get DE genes between conditions not directly listed in the "resultsNames", is the following codes correct? should "batch_2_vs_1" be included int the contrast?##for example, DE between C and F, controlling for batch effect;res.4 <- results(dds, contrast = list("condition_C_vs_A", "condition_F_vs_A")
Best, Chunxuan

	[[alternative HTML version deleted]]

More information about the Bioconductor mailing list