RT-PCR Kit's Validation - (Jul/15/2015 )
Hi all, I need help to understand what's the best way to validate different kits for a qualitative analysis with Real Time PCR. Every kit has values to set up the instrument (n°of cycles, temperatures etc), primers and all the reagents we need.
What I would like to know is which type of method do I need to use to write a validation protocol? I mean..how can I set up number of samples, spikes and positive and negative control? which the minimum number of factors do I need to consider to write a protocol? thank you
It depends on what you mean by "validate". This is a word that many, particularly in academia misuse which might well be a source of confusion to you.
It looks like what you are doing is a comparison of methods and not validating any particular method. Assuming that that is true then what to do is run each method as per each kit's instructions.
Where the comparison comes in is that you need to lock everything you can down to the same conditions. So, the same samples, the same number of samples, run at the same time and by the same person (obviously this is not likely to be possible, but rationalize it and record what you did do).
It looks like what you are doing is comparing each method (or the results gathered by each method) so don't worry that the methods themselves are different.
You should run as many different sample types as you would do on the system you will select and run several runs on each. The exact numbers you need are difficult to say without seeing what you are attempting but a good ballpark would be 3 to 5 separate runs on each instrument each with at least 15 samples each (say 5 samples in triplicate so that you can get an idea of the range of your results within each method). The samples should consist of a number of 'known' samples (samples of known composition and concentration and cover all the concentrations suitable for the methods It's vital that you think through what you need to get out of it - is a particular DNA concentration to be investigated? Will samples often be dirty or come from a 'problem matrix' and then cover this.
Write out what you plan and then ask someone to review it. It's always best to discuss validation plans as it's easy to miss something.
tl:dr In essence a validation is proving that the kit does what it says it does consistently and is suitable for your purposes. Aim with that in mind and you shouldn't go far wrong.