I am going to use PCR-RFLP method for identification purpose. Is it necessary to have replications in the experience? Either in different clones or in amplification or in digestion?
it pretty much depends on your object of study! If you are using human sample for RFLP-identification you will have to run controls to make sure you don't mess up your sample with your own DNA.
If the RFLP is running well, you don´t need a replication. But I agree with Kai, if you use it for identification you need proper controls. I have done it for years in thousands of samples and sometime we had cases of incomplete digest or star activity making a genotyping impossible.
If this is the first time you are doing RFLP, then you must ensure that you have positive samples and standardize your test with these samples. Once you are confident that your reaction conditions are appropriate and the RFLP is working the way it is designed, you can proceed to your actual samples.
Your can test your sample just once, provided the RFLP product is clearly visible and the positive sample has worked the way it is supposed to.
So, one sample, one DNA extraction, followed by PCR and then RFLP, just once!