Entering edit mode
7.7 years ago
EVR
▴
610
HI
I have a RNAseq data of a time course experiment for 3 different time points, T1, T2 ,T3. For every time point, there are samples from 4 different conditions and each sample has triplicates. In total every time point has 12 samples(including replicates).
The extracted mRNA were processed in 3 different batches. Is there any chance, there could be some batch effects on my data as it processed in 3 different batches. If yes, how to find out the batch effect and remove them.
Kindly guide me.
If you use something like edgeR or limma, you can include a bacth effect in your model.
I am using DESeq2 and also without knowing the whether there is batch effect or not how can I include batch effect in edgeR or limma.
You can include batch effect in your model in DESeq2 as well
In the user's guide it is clearly explained how to include a batch effect...
DESeq2 will certainly also have this option (but I never used DESeq2).
Thanks for your suggestion. But how to find whether there is. Batch effect in my rnaseq data?
Check with MDSplot (from egdeR) if there is an indication of a batch effect.
Besides, it won't hurt to include the batch effect in the model if you expect it.
Don't forget that the analysis of batch effect heavily depends on the experimental design. If you have not make a right design, it will be impossible to study the batch effect
I second that!
If you have let's say time point 1 in batch 1, time point 2 in batch 2, etc. Then of course it doesn't make sense to include batch effect in your analysis.