Hi, I am doing a relative quantification of genes (A and B) from same family.I am using Actin as calibrator. I have primers specific for gene A and primers specific for gene B. If I know these two genes has the same function and I would like to know A+B expression, is it correct to sum dCq from A + B or do I need to design new primers to amplify both at the same time if I really want to know the expression of both genes together? For example: dCq for gene A is 10 and for gene B is 5. Can I assume A+B dCq is going to be 15? If it not true why it is not that simple? Thank you.
[Edit: I had A and B reversed in my original answer here: now corrected below]:
Since the kinetic of each different target qPCReaction differs from minimally to greatly (e.g., efficiency of amplification, dynamic range, onset of Poisson noise, etc.), they should be considered as separate and distinct indicators of single target behaviors.
But, for the sake of example, even if we consider your premise that A and B can be summed in some way as equal targets:
Realize that dCq values are log scale entities/values. For the sake of simplicity, assuming 100% efficiency (EAMP = 2) for your targets: A, B and Actin in your example: Consider Actin in your case to be the baseline reference gene Cq that is subtracted from the Cq values for A and B. First note that (for B) a dCq of 5 from the Actin baseline Cq indicates 32 times more target than (for A) a dCq of 10. And a dCq of 15 represents 1024 times less target than in B. So you see the problem right there. Relative quantity involving a housekeeping (HK) or reference gene (RG) (e.g., Actin here), in the simplest sense, can be expressed as EAMP(-dCq) and using "2" for EAMP (which is seldom the case for all targets all of the time) allows you to see this.
A dCq of 5 indicates an actin normalized relative quantity of: 0.03125
A dCq of 10 indicates an actin normalized relative quantity of: 0.000976563
A dCq of 15 indicates an actin normalized relative quantity of: 0.000030517578125
So, here you see that if you "add" 0.3125 + 0.000976563, this = 0.0322265625 (as also mentioned by Michael above)
Next you see that using a dCq of 15 to represent the result is nonsensical since 0.000030517578125 does not equal 0.0322265625, in fact it represents a relative actin normalized quantity that is 1056X less than 0.0322265625 - so this is a dangerous assumption to say the least.
The operation you are probably seeking is the one highlighted in bold above - and some of the assumptions there are OK, but for the most part, this is an unorthodox approach to processing qPCR data (although orthodoxy has its own pit-falls as well).
With blinders on (meaning: assuming that all of your target reactions A, B, and Actin, are 100% efficient, and that the actin normalized relative linear, normal scale quantities are indeed additive), as long as you state what your assumptions are from the outset, and that your are working with the log transformed linear scale (normal scale) Actin normalized relative quantity values to accomplish your ends, you could, in theory be relatively OK.
Another point: the definition of a "calibrator" is a generally a control sample of some kind wherein target and housekeeper (or reference gene) are both measured as well, and then that HK or RG normalized target value is what all other HK or RG normalized treatment group target values are divided by.
Measuring the efficiency of each of your target and HK or RG reactions is very important as it can greatly impact your final results given the underlying log base nature of the qPCR gymnastic.
This all being said, and going even more in an unorthodox direction ... if you do measure both signals (A and B) in the same reaction to get a composite Cq value and it is not important for you to know the proportion of A to B, and your Actin Cq value is generated from the exact same sample that A and B are being measured, your hunch of getting at a general picture of 'functional transcript abundance' then would be similar to measuring multiple exons in the same reaction. Again, this is assuming ~100% efficiency for all players. Additionally, if you are using SYBR Green based qPCR, you could theoretically see both of your amplicon products (A and B) during melt curve analysis at the end of the runs as well.
In short, your original approach of adding dCq for A to the dCq for B, was actually giving you the ~opposite/inverse result of what you are seeking...
If you are using the delta Ct method, and assuming 100% amplification efficiency (an imperfect, but common assumption), you would conclude that A is expressed at 1/2^10 compared to actin and B is 1/2^5 compared to actin within the same sample. The amount of the two transcripts combined relative to actin would be the sum of these values, which would be a miniscule bit less than 1/2^5. Usually, this approach is used to compare the same gene relative to a calibrator between samples (delta delta Ct method).
[Edit: I had A and B reversed in my original answer here: now corrected below]:
Since the kinetic of each different target qPCReaction differs from minimally to greatly (e.g., efficiency of amplification, dynamic range, onset of Poisson noise, etc.), they should be considered as separate and distinct indicators of single target behaviors.
But, for the sake of example, even if we consider your premise that A and B can be summed in some way as equal targets:
Realize that dCq values are log scale entities/values. For the sake of simplicity, assuming 100% efficiency (EAMP = 2) for your targets: A, B and Actin in your example: Consider Actin in your case to be the baseline reference gene Cq that is subtracted from the Cq values for A and B. First note that (for B) a dCq of 5 from the Actin baseline Cq indicates 32 times more target than (for A) a dCq of 10. And a dCq of 15 represents 1024 times less target than in B. So you see the problem right there. Relative quantity involving a housekeeping (HK) or reference gene (RG) (e.g., Actin here), in the simplest sense, can be expressed as EAMP(-dCq) and using "2" for EAMP (which is seldom the case for all targets all of the time) allows you to see this.
A dCq of 5 indicates an actin normalized relative quantity of: 0.03125
A dCq of 10 indicates an actin normalized relative quantity of: 0.000976563
A dCq of 15 indicates an actin normalized relative quantity of: 0.000030517578125
So, here you see that if you "add" 0.3125 + 0.000976563, this = 0.0322265625 (as also mentioned by Michael above)
Next you see that using a dCq of 15 to represent the result is nonsensical since 0.000030517578125 does not equal 0.0322265625, in fact it represents a relative actin normalized quantity that is 1056X less than 0.0322265625 - so this is a dangerous assumption to say the least.
The operation you are probably seeking is the one highlighted in bold above - and some of the assumptions there are OK, but for the most part, this is an unorthodox approach to processing qPCR data (although orthodoxy has its own pit-falls as well).
With blinders on (meaning: assuming that all of your target reactions A, B, and Actin, are 100% efficient, and that the actin normalized relative linear, normal scale quantities are indeed additive), as long as you state what your assumptions are from the outset, and that your are working with the log transformed linear scale (normal scale) Actin normalized relative quantity values to accomplish your ends, you could, in theory be relatively OK.
Another point: the definition of a "calibrator" is a generally a control sample of some kind wherein target and housekeeper (or reference gene) are both measured as well, and then that HK or RG normalized target value is what all other HK or RG normalized treatment group target values are divided by.
Measuring the efficiency of each of your target and HK or RG reactions is very important as it can greatly impact your final results given the underlying log base nature of the qPCR gymnastic.
This all being said, and going even more in an unorthodox direction ... if you do measure both signals (A and B) in the same reaction to get a composite Cq value and it is not important for you to know the proportion of A to B, and your Actin Cq value is generated from the exact same sample that A and B are being measured, your hunch of getting at a general picture of 'functional transcript abundance' then would be similar to measuring multiple exons in the same reaction. Again, this is assuming ~100% efficiency for all players. Additionally, if you are using SYBR Green based qPCR, you could theoretically see both of your amplicon products (A and B) during melt curve analysis at the end of the runs as well.
In short, your original approach of adding dCq for A to the dCq for B, was actually giving you the ~opposite/inverse result of what you are seeking...
- Badges
- Science topic