be sure before dephosphorylating that your plasmid is really 100% digested, because dephosphorylation obviously won't act on uncut DNA, furthermore uncut DNA is more problematic than religated DNA because supercoiled plasmid transforms more efficiently.
i agree with Olivier, that this is the most seious drawback. I think that as in many experiments like this one, that alkaline phosphatase will not work 100% so there is always a small danger that you might religate digested plasmid. Nevertheless if you treat your digested plasmid with alkaline phosphatase only a small protion can be religated and the majority of your plasmids should be avaiable for introducing the desired piece of DNA
Hi, there is another possibility, that in my case works better than dephosphorilation and it is easier. You need to digest your ligation with an enzyme that cuts the plasmid but not the one with your insert. Of course you can not always do that, but it is a good method when you have a MCS and betweem the restriction sites that you are using for your cloning there are others that are not present in your insert, for example.
Either you can use Alkaline Phosphotase to dephosphorlate the plasmid before ligation or you may use double digestion strategy (using two different restriction sites to open the plasmid) for your ligation.
Phosphatase treatment is good. The main issue with a one-enzyme digest will be that you can not know which way the insert will. I would suggest double-digest, it's as easy and you'll have a better chance to get the product you are looking for.
You could run some cut DNA on a gel, make sure to purify the linear band (to ensure complete digestion) then go a head and dephosphorylate. That will take care of both incomplete digestion and religation of empty vector.
Make sure to run a good amount of DNA in the gel and use TAE not TBE for the buffer.