I usually perform qPCR after stimulating cells, so for every donor I normalized the mRNA expression to the unstimulated sample. But now I have to compare mRNA expression among mice infected with different vectors. As I have many genes to compare, I usually put 2 mice in every plate. Which is the best way to normalize results and compare data from individual mice? Is housekeeping enough or do we need some standard?
Hi Francesco, have a look here http://pathmicro.med.sc.edu/pcr/realtime-home.htm
Generally people run one gene per plate with all the different animals (n), rather than all genes of one mouse in one plate. It's easier and more accurate to compare samples in one plate and that's what you usually are interested anyway - the changes of one gene across a sample population. For the housekeeping genes - the more the better. How you evaluate your data, see the link. - Andreas
Hi Francesco, have a look here http://pathmicro.med.sc.edu/pcr/realtime-home.htm
Generally people run one gene per plate with all the different animals (n), rather than all genes of one mouse in one plate. It's easier and more accurate to compare samples in one plate and that's what you usually are interested anyway - the changes of one gene across a sample population. For the housekeeping genes - the more the better. How you evaluate your data, see the link. - Andreas
I agree - if possible, run all the samples that you want to compare directly on one plate as the variability between plates can be quite high. And yes, the more housekeeping genes you use, the better - good luck! - Nikki
Just one word again to housekeeping genes. It can be quite challengin to find a gene that doesn't change upon your treatment/conditions. I think a very common problem is that people just assume GAPDH or tubulin or similar genes are stable, but they might actually be dysregulated in your system. Better to invest some plates to make sure your normalisation is stable, because all your data will be based on this. -Andreas
Better to add all mice cDNA in the same plate, as stated above. In order of comparison, if u dont want to do absolute quantification, u could use an internal calibrator made of pooled cDNA from all mice. Hence, u ll have arbitrary scale but everything normalized to the same "control
sample", thus comparable between them.
In addition to housekeeping (in my conditions I used b-actin or 18S) you could also realize a pool of RNA from some unstimulated samples and then a standard curve with different dilution of this to be used in all test.
Hello. 1) It would be wise to put your reference genes to the test, so that you confirm if they are stably expressed (or with low variation) across your treatments/samples. Heads up! the more is not the best here. You can test that with GeNorm or Normfinder software, I believe that Genex also does that, but I am not sure. (best keeper also does something similar). 2) As for the comparison among samples, I agree with the fellows above: Depending on what your questions are, usually the more effective set up is putting as much samples as you can in the same plate, so they all will have the same technical variation effects; and split the genes in several plates. Hellemans; vandesompele et al (2007) call that "sample maximization set up". One last thing: in that same paper you can find a sugestion about how to normalize variation amongst plates. Hope it helps.
- Badges
- Science topic
- Similar topics
- Biological Science
- Molecular Biology
- Biomolecules
- Nucleic Acids
- RNA