Author: Wang, Jun; Zheng, Xiaoqi
Title: WSE, a new sequence distance measure based on word frequencies Cord-id: d6wcgkln Document date: 2008_9_30
ID: d6wcgkln
Snippet: Abstract In this article, we present a new distance metric, the Weighted Sequence Entropy (WSE), based on the short word composition of biological sequences. As a revision of the classical relative entropy (RE), our metric (1) works equivalently with RE in the case of small k, (2) avoids the degeneracy when some word types are absent in one sequence but not in the other. Experiments on 25 viruses including SARS-CoVs show that our method and RE give exactly the same phylogenetic tree when word le
Document: Abstract In this article, we present a new distance metric, the Weighted Sequence Entropy (WSE), based on the short word composition of biological sequences. As a revision of the classical relative entropy (RE), our metric (1) works equivalently with RE in the case of small k, (2) avoids the degeneracy when some word types are absent in one sequence but not in the other. Experiments on 25 viruses including SARS-CoVs show that our method and RE give exactly the same phylogenetic tree when word length k ⩽ 3 . When k > 3 , our method still works and gets convergent phylogenetic topology but the RE gives degenerate results.
Search related documents:
Co phrase search for related documents, hyperlinks ordered by date