<<
Return to qPCR page
Reference Gene qPCR
Goals
- Determine which of your reference genes you are going to normalize to.
Why am I doing this?
Reference genes need to be stable across your control and experimental samples in order to be useful. If expression between the two differs, you will be normalizing to two different values and that is worthless. This is a common omission. Reference genes must be validated.
Typical plate setup for three candidate reference genes:
| 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 |
A
|
C1BR1
|
C1BR1
|
C1BR1
|
C1BR1
|
C1BR1
|
C1BR1
|
C1BR1
|
C1BR1
|
C1BR1
|
B
|
C1BR2
|
C1BR2
|
C1BR2
|
C1BR2
|
C1BR2
|
C1BR2
|
C1BR2
|
C1BR2
|
C1BR2
|
C
|
C1BR3
|
C1BR3
|
C1BR3
|
C1BR3
|
C1BR3
|
C1BR3
|
C1BR3
|
C1BR3
|
C1BR3
|
D
|
C1BR4
|
C1BR4
|
C1BR4
|
C1BR4
|
C1BR4
|
C1BR4
|
C1BR4
|
C1BR4
|
C1BR4
|
E
|
C1BR5
|
C1BR5
|
C1BR5
|
C1BR5
|
C1BR5
|
C1BR5
|
C1BR5
|
C1BR5
|
C1BR5
|
F
|
C2BR1
|
C2BR1
|
C2BR1
|
C2BR1
|
C2BR1
|
C2BR1
|
C2BR1
|
C2BR1
|
C2BR1
|
G
|
C2BR2
|
C2BR2
|
C2BR2
|
C2BR2
|
C2BR2
|
C2BR2
|
C2BR2
|
C2BR2
|
C2BR2
|
H
|
C2BR3
|
C2BR3
|
C2BR3
|
C2BR3
|
C2BR3
|
C2BR3
|
C2BR3
|
C2BR3
|
C2BR3
|
I
|
C2BR4
|
C2BR4
|
C2BR4
|
C2BR4
|
C2BR4
|
C2BR4
|
C2BR4
|
C2BR4
|
C2BR4
|
J
|
C2BR5
|
C2BR5
|
C2BR5
|
C2BR5
|
C2BR5
|
C2BR5
|
C2BR5
|
C2BR5
|
C2BR5
|
| | | | | | | | | |
|
ref gene 1
|
|
|
ref gene 2
|
|
|
ref gene 3
|
|
|
| | | | | | | | | |
|
C1 = CONDITION 1
| | | | | | |
|
C2 = CONDITION 2
| | | | | | |
|
BR# = BIOLOGICAL REPLICATE #
| | | | |
Conditions
What you are looking for
- Technical replicates with a standard deviation below 0.2 (this is arbitrary and most of your replicates will be below 0.1. If you do enough qPCR, you will eventually become obsessed with how low you can get this number).
- At least 2 reference primer sets that show no significant difference between control and experimental conditions. The standard deviation of the Cqs for all biological replicates should be low. (Preferrably Less than 0.5)
- If you are seeing differences between replicates or perhaps conditions, you may be asking the question “how do I know if there’s really a difference, it could be something else like loading or RT-PCR efficiency? I am not controlling for any of these by normalizing to anything!”. The answer is if you prepared good quality RNA and loaded exactly the same amount of RNA into a well-prepared reverse transcription, there should be very little (less than 1 Cq) difference between biological replicates of the same condition, assuming that condition itself is reproducible. If the biological variation is truly large between replicates, you’ll have to pick the best you can.
<<
Return to qPCR page
Topic revision: r1 - 2019-03-21 - 21:51:06 - Main.KateElston