Our study comprises of PCR analysis of DNA isolated from fecal samples at two time points (at the baseline and after intervention). We have carried out PCR and have obtained the Ct values. However, we did not use a reference gene and hence cannot go for delta Ct value. Is it possible to apply t-test or any other statistical test to these two sets of Ct values directly?
Of course it's technically possible to apply T-test on any two sets of Ct values. But the problem is you would surely be looking for differential gene regulation relative quantification which is only possible when you normalize the Ct values with a suitable reference gene.
How about you amplify your gene through PCR several times gene clean it (I hope it's a good sized amplicon), quantify the PCR Product make serial dilutions and go for Absolute quantification of your Ct values through Regression analysis.
Still have no idea how will Peer reviewers respond to this technique.
I hope it helps.
Dear colleague:
Comparing just Cts does not seems completly correct. I agree with the previous answer, you can prepare a standard curve and perform an absolute quantification. Also you can try adding to your cDNA batch a known quantity of a PCR amplicon that you can use as a external reference transcript that you can use as a normalizer.
Best regards
- it is NOT technically possible
- You need to extact actual relative gene expression data by the Livak method and then having found factorial differences in expression between treated and control sample you then plus those numbers into a t test
- To do that with un normalised raw Ct values is completely non sensical !!
- Find attached worked examples on how to compute relative expression by the Livak method
- Also see many answers I have posted on this subject:
- https://www.researchgate.net/post/How_can_I_calculate_gene_expression_levels_with_RT-PCR#view=54994b14cf57d76a568b45a3
Laurence Stuart Dawkins-Hall Felipe Ignacio Aguayo Shoaib Ahmed
Thank you all very much for the replies. I forgot to mention that the all the DNA samples were normalized to 1ng per microlitre before carrying out the PCR. In that case is it okay to go ahead with t-test as I have mentioned earlier?
We are working on fecal samples and are looking for specific groups of bacteria in them. Which gene can be used as a reference ?
Thanking you all in advance.
Dear colleague:
I strongly recommend you the addition of an external cDNA (or amplicon) to use as a reference transcript. Thinking in your case, I would use some non at all related cDNA, such as Arabidopsis.
Wish you luck!!!!!!!
- I still cannot see how that would work because differences in gene expression from one ct value to the next are exponential
- In other words if you had say a difference of 3 intergers in terms of your ct value then your actual difference in gene expression would be 2^(3); that is 8 fold not 3 fold
- So no you still need to convert ct differences to actual gene expression
- Otherwise why bother with the Livak method ?
Hi only word of caution is central limit theorem of statistics,which states that if ur sample size small then u can not assume it normal distribution,so basically best way to represent ur data is find mean mode median if they r overlapping proceed to the t test,t distribution adjust itself according to degree of freedom u r giving which is ur smoke size -1,otherwise u can always go ahead to do std curve plot ur ct on it to get estimate of absolute quantity which u will plot and compare with t statistic ur control.
Hope that answers,
- Badges
- Science topic