I'm using edgeR for differential expression analysis of RNASeq data (counts downloaded from TCGA).
The workflow I'm using follows the GLM functionality of edgeR, as described in the vignette for edgeR.
dge <- DGEList(...) design <- model.matrix(...) # non-singular y <- estimateGLMCommonDisp(dge, design) y <- estimateGLMTrendedDisp(y, design) y <- estimateTagwiseDisp(y, design) fit <- glmFit(y, design) lrt <- glmLRT(fit, coef = ncol(design) + c(-1,0))
Fitting the model, as above, can take ages on our server - seemingly because the dataset is pretty large, the design matrix is pretty large etc etc AND, only a single processor is used by edgeR.
I can't find any options within edgeR that can take advantage of parallel processing and was wondering whether any of the alternative RNASEq expression packages, such as DESeq, had this capability
All the best