I am having a problem to analyze the differentially expressed genes by using cuffdiff , because besides the two treatments I also have samples of different sex, and I would like to distinguish beteween them ?
Is it possible to run a 2 step approach using cuffdiff (first with sex effect and repeat with treatemnt) ? If so , which files should I load in the program?
Which alternative programs should I use after aligment to analyze this model with two effects (treatment and sex).
Thanks
•
written
by
TCC •
1 answer
I don't believe that you can include a second variable for cuffdiff.
However, you can do so for edgeR, limma-voom, DESeq2, and even a regular ANOVA or linear regression.
In fact, you might find those strategies advantageous over cuffdiff (even though cufflinks can be useful for reference guided de novo assemblies):
http://cdwscience.blogspot.com/2019/11/requiring-at-least-some-methods-testing.html
In terms of the relevant commands:
edgeR:
design = model.matrix(~var1 + var2)
fit = glmFit(y, design)
lrt = glmLRT(fit, coef=2)
test.pvalue = lrt$table$PValue
(you can also add y = estimateGLMRobustDisp(y, design) for edgeR-robust)
limma-voom:
design = model.matrix(~var1 + var2)
v = voom(y,design,plot=TRUE)
fit = lmFit(v,design)
fit = eBayes(fit)
pvalue.mat = data.frame(fit$p.value)
DESeq2:
dds = DESeqDataSetFromMatrix(countData = deg.counts, colData = colData, design = ~ var1 + var2)
You can then calculate a p-value via Wald Test (dds = DESeq(dds) followed by res = results(dds, contrast = c("var1", trt.group, other.groups))) or Likelihood Ratio Test (dds = DESeq(dds, test="LRT", reduced = ~ var2) followed by res = results(dds)).
Log in to answer this question.
-
RNA-Seq with more conditions
written by AriBo •
-
How to perform DESEQ2 analysis when there are three treatments (one of them being a reference)?
written by mrj •
-
Statistical analysis for proteomic data with more than one effect in the model
written by TCC •
-
RNA-Seq results comparison
written by Spacebio
-
how to deal with batch effect in TCGA RNA-seq dataset
written by tujuchuanli
-
Analyzing RNA-seq without replicates
written by shawn.w.foley
-
Why does cuffdiff test differential isoform expression for genes with only one annotated transcript?
written by unsupervised_learner •
-
[DESeq2 multiple treatments vs. 1-by-1] Inconsistant p-adj
written by madkitty
-
GLM design for RNAseq analysis using DESeq
written by illinois.ks
-
How To Transform Microarray Data To Adjust For Batch Effects
written by David Quigley
Do you realize that cuffdiff is a tool for differential transcript rather than gene level analysis? Better use the tools Charles Warden mentions below. Only recommendation I have when using edgeR is to use the QLF pipeline (check edgeR manual) which is the current recommendation of the authors, claiming to be slightly superior to the LRT pipeline.