We are doing site-saturation mutagenesis of LeuDH. But when we are mutating Q to G, the result of the sequence is always wrong, only a small fragment is the same to our templet plasmid. I have done it for three times, the results are the same. However, other mutation were all successful for the first time.