How to deal with stronger effect by donor than treatment
1
0
Entering edit mode
Pei • 0
@e9de1a10
Last seen 2.3 years ago
United States

Hello everyone:

I have 3 pairs of samples (treatment + control) from 3 donors.

When perform PCA or MDS, I found samples were separated into 3 clusters, each corresponding to 1 donor. Consistently, number of DEGs were few between treatment and control.

Is it possible, or reasonable, to remove such donor effect before further analysis?

Thanks!

Pei

/* Here are some related codes:

donor = factor(coldata$Patient)

Treat = factor(coldata$Treatment, levels=c("Control","One_important_gene")))

design = model.matrix(~donor+Treat)

y = DGEList(counts=matrix_of_raw_read_count)

keep = rowSums(cpm(y) > 2.0) >= 3

y = y[keep,,keep.lib.sizes=FALSE]

y = calcNormFactors(y)

plotMDS(y)

y = estimateDisp(y,design,robust=TRUE)

fit = glmFit(y,design,robust=TRUE)

lrt = glmLRT(fit,coef="One_important_gene")

*/

MDS RNA-Seq • 1.2k views
ADD COMMENT
2
Entering edit mode
@gordon-smyth
Last seen 1 hour ago
WEHI, Melbourne, Australia

The donor effect has already been completely accounted for in the differential expression analysis. That's why we recomment a paired analysis for this sort of experiment with donor in the design matrix. There is no need for you to do anything else

The fact that the donor effect is larger than the treatment effect does not prevent you from getting DEGs. What does prevent DEGs would be if the treatment effects are inconsistent between the donors. If the treatment effect is small and inconsistent between the donors then it is correct and appropriate for the statistical method to return few DEGs.

ADD COMMENT
0
Entering edit mode

Thank you Gordon!

ADD REPLY

Login before adding your answer.

Traffic: 699 users visited in the last hour
Help About
FAQ
Access RSS
API
Stats

Use of this site constitutes acceptance of our User Agreement and Privacy Policy.

Powered by the version 2.3.6