Tm variation - (Jul/09/2012 )

Hi guys,
I have a question...
I am analyzing a Tm curve, and for an amplicon size of 171bp.
I have got a Tm variation within replicates and among different samples from 83 to 85.
Could that be possible?? What does it mean??
What should be the maximum accepted Tm variation, 1, 2 , 3 degrees??
When I say this 2 products are differents??

THanks a lot in advance!

That depends a lot on the temperature homogeneity of your cycler. You could test it on a product you are sure is specific, make many replicates and see what variation you get.
I usualy look more on the shape of the curve since the Tm calculating algorithm could have a variation of it's own. Mostly nonspecific products have two clearly distinct peaks or the main peak has a "sidepeak" or is unusually wide. Also changes within sequence in addition to product length can cause Tm shifts, so there can be mutation or polymorphism causing some variation, but these changes are usually too subtle to be detected by SYBR green chemistry.

Other thing you have to consider is that you usually expect your desired produc or your desired product plus something else. It would be very unusual to have a nonspecific product melting very close to your desired while not having desired product at all, which would cause widepeak or sidepeak (however I can of course imagine cases where this could really happen). So if you see a clear sharp distinct peaks with similar heights just a bit apart it's most likely still the same product, though you should check the homogeneity as I mentioned earlier.

If you're still not sure whether they're not different products that differ only slightly in length, you can run your products on polyacryamide gel to check teh length.