In the era of information explosion, biological data is being produced at an alarming rate. With the amount of data currently available, computational methods have become an integral part of sequence alignment. In protain homology analysis and sequence alignments, the demand for the distinguish of specific gene mutation types has been increased rapidly in recent years. In this paper, we presents an efficient algorithm to distinguish gene mutation types. For the input DNA pairs, the algorithm outputs the mutation type, the location of the mutation, and the specific base changes conditions. At last, the simulation result shows the effectiveness of the algorithm.