I have six samples, 3 replicates of "Naive CD8 T cells" and 3 replicates of "DMSO-activated T cells". I am doing an RNA-seq differential gene expression analysis to compare the expression levels of genes between Naive CD8+ cells and activated CD8+ cells, and produced this PCA plot in R. Can you interpret what this plot is saying about my samples?
Hi Lilia,
first of all the PCA mainly shows what you would expect: the principle component 1 discriminates between the conditions (naiive/activated). Furthermore, samples are separated by some measure in PC2 in which it seems like there is a connection between one replicate each of the two conditions. The easiest explanation would be some "systematic bias" in the samples, for example the lower two points are from donor 1, the middle from donor 2 and the upper ones from donor 3.
PC2 explains 19% of the variance, which means it is a much smaller contributor compared to PC1 (57%).
A handy explanation of PCA plots can be found here: https://setosa.io/ev/principal-component-analysis/
Hope this helps!
Alexander
Yes, you see a clear effect of DMSO activation, as these two groups separate nicely on PC1. Actually, there is nothing more this biplot shows. However, it would be interesting to color or shape or label the 3 replicates to see if it's the replication that sorts the pairs across PC2 (as Alexander guessed). That's easy to do. If you see this, a paired analysis would be be advisable to investigate the DMSO effect.
Lilia Olmstead
If I would be you, before interpreting, I would have asked is it correct? Because by seeing your graph, I would say something is not right there.
- Badges
- Science topic
- Similar topics
- Mathematics
- Statistics
- Statistical Software
- R