A novel estimation algorithm of evolutionary distance based on the word frequency variation of sequence segment

摘要

Based on the model of evolution, studies on the genome sequence of organisms use alignment method to calculate evolutionary distances between organisms. However, there is clear disadvantages in that it requires high complexity of the whole algorithm and necessity of perfect sequence assembly in the conventional studies. Therefore, in recent years, alignment-free methods for analyzing the relationship between sequences by using fragmentary information such as word frequency or common substring of each sequence without progressing alignment for the entire sequence are being researched. Nevertheless, the alignment-free method has not been widely used yet. This is because the alignment free methods so far have presented algorithms which are solely experimental rather than mathematical. Therefore, we propose a mathematical estimation method of the evolutionary distance based on the model of evolution and prove that it shows superior performance compared with various alignment-free methods.