Study on the Relation between Virus and Host Cell by Alignment-Free Sequence Comparison

摘要

With the development of the human genome project, scientists have obtained a large number of biological sequences which needed to be processed and analyzed. In this paper, we characterized the biological relationship between virus and host cell by using alignment-free sequence comparison statistics such as, D2S, D2*, Hao, Eu and Ch. By using sequence alignment-free comparison, we acquired sequence distributions of K-tuple in the tested species, and here K-tuple in biology is a short sequence of genes with a length of K. We used the ROC curves to analyze the data of DNA sequences of virus and host cell to find an optimal statistic method. The ROC curves show that when K is small, all of the statistics can better react the biological relationship of virus and host cell in the genetic similarity, but with K gradually increasing, the statistics D2S and D2* have been proved to be almost invariant and well in reflecting the biological relationship between virus and host cell in the genetic similarity, at the same time the other statistics reduced, which confirmed our conjecture, so we can consider the statistics D2S and D2* as effective alignment-free sequence comparison methods for the relation between virus and host cell.