Hi Folks,
I am currently working on 16 bulk RNAseq samples. I did some basic research on statistical testing of differential expression and figured out two R packages to edgeR and DESeq2.
The 16 samples are actually paired samples data from before treatment and after treatment from 8 patients. I am interested in finding DE genes within paired samples for instance (P1_AT vs P1_BT). This is the design matrix
Sample Patient Treatment
P1_BT P1 BT
P1_AT P1 AT
P2_BT P2 BT
P2_AT P2 AT
........
P8_BT P8 BT
P8_AT P8 AT
So I have created my design matrix like section 4.1
- design <- model.matrix(~ Patient+Treatment)
EdgeR manual, my experiment design is exactly similar to section 4.1 of edgeR manual.
4.1 RNA-Seq of oral carcinomas vs matched normaltissue
Patient <- factor(c(8,8,33,33,51,51))
Tissue <- factor(c("N","T","N","T","N","T"))
data.frame(Sample=colnames(y),Patient,Tissue)
Sample Patient Tissue
8N 8 N
8T 8 T
33N 33 N
33T 33 T
51N 51 N
51T 51 T
design <- model.matrix(~Patient+Tissue)
rownames(design) <- colnames(y)
design
(Intercept) Patient33 Patient51 TissueT
8N 1 0 0 0
8T 1 0 0 1
33N 1 1 0 0
33T 1 1 0 1
51N 1 0 1 0
51T 1 0 1 1
fit <- glmFit(y, design)
lrt <- glmLRT(fit)
topTags(lrt)
Coefficient: TissueT
RefSeqID Symbol NbrOfExons logFC logCPM LR PValue FDR
5737 NM_001039585 PTGFR 4 -5.18 4.74 98.7 2.97e-23 3.12e-19
5744 NM_002820 PTHLH 4 3.97 6.21 92.2 8.00e-22 4.21e-18
3479 NM_001111283 IGF1 5 -3.99 5.71 86.5 1.38e-20 4.85e-17
colnames(design)[1] "(Intercept)" "Patient33" "Patient51" "TissueT"
But I am interested in getting 1) P1_AT vs P1_BT, 2) P2_AT vs P2_BT ... 8) P8_AT vs P8_BT
To find genes with baseline differences between the drug and the placebo at 0 hours:
qlf <- glmQLFTest(fit, contrast=my.contrasts[,"DrugvsPlacebo.0h"])
my.contrasts <- makeContrasts(
+ Drug.1vs0 = Drug.1h-Drug.0h,
+ Drug.2vs0 = Drug.2h-Drug.0h,
+ Placebo.1vs0 = Placebo.1h-Placebo.0h,
+ Placebo.2vs0 = Placebo.2h-Placebo.0h,
+ DrugvsPlacebo.0h = Drug.0h-Placebo.0h,
+ DrugvsPlacebo.1h = (Drug.1h-Drug.0h)-(Placebo.1h-Placebo.0h),
+ DrugvsPlacebo.2h = (Drug.2h-Drug.0h)-(Placebo.2h-Placebo.0h),
+ levels=design)
3.3.1 Defining each treatment combination as a group
The design matrix is different for section 4.1 and section 3.3.1. My experiment similar to section 4.1, so I have used that design matrix. But I want the results of section 3.3.1 within samples comparison.
Any comments are appreciated.
