Entering edit mode
2.4 years ago
Azade
▴
20
Hi all, Are these samples good enough for DEA analysis? Because I think the tumor and normal samples are not well separated.
Here is my code:
NormalizedCounts.scaled <- t(scale(t(NormalizedCounts) , scale = F))
pc <- prcomp(NormalizedCounts.scaled)
pcr <- data.frame(pc$rotation[,1:3] , Group = gr)
ggplot(pcr, aes(PC1 , PC2 , color = Group)) + geom_point()
Thanks for any help.
This is an awesome separation for clinical data, you will often just see a mess of points with no separation at all because of all the potential confounders in human specimen that obscure the actual clinical effect you're after. Go ahead, plug into your favourit DE software and see what comes out.
Thanks for your positive feedback.