Hi there,
I want to use ddCt to analyze my qPCR results, however I am not exactly sure on how to proceed and when to average the replicates. Also, can we use ddCt when having more than 1 factor (e.g. treatment dose and treatment time)?
For example, I performed a cell culture experiment. I did 3 replicates (one each month during three month). I had two treatment times and three treatment doses (+ untreated control).
For each replicate experiments, cells were treated for two hours, or for 24hours, with a concentration of 0M (control), 10-7M, 10-6M or 10-5M. Cells were harvested, and using qRT-PCR I analyzed the expression of target genes + 3 reference genes.
In this experiment I want to see 2 things :
1. If gene expression is different in treated cells vs untreated control.
2. If gene expression is different according to duration of treatment (2h vs 24h).
I am not sure about how to correctly use ddCt and how to compare my two factors.... Should I first take the 2h experiment and calculated ddCt for treated vs untreated and than do the same for the 24h ? Then how should I compare times (2h vs 24h) using the ddCT ? Also, should I right away average my Ct values of the 3 replicates ?
I've attached my excel file with my ddCT calculations.
I really want to understand the rationale behind it..!
Thanks for your help
This is a classical 2x2 design. Your response variable (dCt values) can be assumed normal ditributed, so you can use standard procedures (2-way ANOVA) to analyze such experiments.
There is a lot you will find if you google "factorial design" or "2x2 design". Here is one hit:
https://crumplab.github.io/statistics/factorial-anova.html
Thank you Jochen Wilhelm , and taking the option where I would like to just keep one time experiment. For example, If I would like to analyze ddCt on only the 2h treatment experiment.
Am I correct in the way I've calculated my ddCt values? In the excel file, I calculated ΔΔCts by subtracting ΔCt CTRL (mean of the 3 reps) to each of the ΔCt treatment (mean of the 3 reps as well) ? Or its not ok to use the same CTRL ?
I have done :
ΔΔCt treatment1 = (ΔCt treatment 1 - ΔCt CTRL)
ΔΔCt treatment2 = (ΔCt treatment 2 - ΔCt CTRL)
ΔΔCt treatment3 = (ΔCt treatment 3 - ΔCt CTRL)
Also, is it okay to directly average the dCt obtained for my three replicates ?
Thank you!
It seems that the treated samples and the controls are independent. It is therefore not meaningful to calculate an individual ddCt for each treatment replicate. Simply average the dCt values of the replicates, and subtract the average dCt for the controls from the average dCt of the treated samples. This is the estimated ddCt. If you want a standard error ora confidence interval for this estimate, I suggest using a stats software, where you can enter the (two groups of) dCt values, and which gives you the estimate plus all statistics (just like you would in the case of a 2x2 design).
Jochen Wilhelm Thank you very much, I am not sure I understand what you mean when you say they are independent...
when you said : ". Simply average the dCt values of the replicates, and subtract the average dCt for the controls from the average dCt of the treated samples. " , I do this for each treatment dose right ?
I've modify my excel file, I think I did it right this time ?
It gives me 4 ddCTs
1. CTRL (which is 0 of course)
2. dose 1 ; 10-7 (ddCt = -0.19)
3. dose 1 ; 10-6 (ddCt = -0.12)
4. dose 1 ; 10-5 (ddCt = -0.09)
The calculations are correct.
I would not calculate a ddCt for CTRL - this makes no sense.
may be this might be helpful for you: Method Excel File to calculate relative gene expression using one h...
- Badges
- Science topic
- Similar topics
- Biological Science
- Genetics
- Gene Expression