I am doing mutation pcr using quickchange mutation. While using pfu which has proof reading activity, I am getting wrong mutated product for example getting more g or duplication or mutation at wrong nucleotide region.
Non- specific binding is the problem. There are too many binding configurations with similar annealing temperatures. You could try to use different primers and/or temperatures and/or buffer/salt concentrations. I have had some luck using higher temperatures with 10% DMSO to start out with.
Although there are some explanations of your strange results but there is no easier way for you to sort them out using the quickchange protocol. The main problem is its low efficiency of PCR amplification. There is a modified method that explains the problems of the QuickChange and reports a modified method with high efficiency (http://bmcbiotechnol.biomedcentral.com/articles/10.1186/1472-6750-8-91). It might be much quicker for you to solve your problems by changing your primers as described in the modified protocol. If you like you can follow the attached lab protocol based on that published method.