基于基因本体的相似度计算方法  

Computation method for semantic similarity based on gene ontology

作　　者：荣河江 王亚东[1] RONG Hejiang;Wang Yadong(School of Computer Science and Technology,Harbin Institute of Technology,Harbin 150001,China)

机构地区：[1]哈尔滨工业大学计算机科学与技术学院,哈尔滨150001

出　　处：《智能计算机与应用》2019年第1期108-113,118,共7页Intelligent Computer and Applications

摘　　要：对基因或基因产物功能上的相似性的研究是生物信息学关注的热点。基因本体数据描述了基因和基因产物的属性,基于基因本体术语相似度的研究对基因功能的分析、比较和预测具有重要意义。本文研究了基于基因本体的语义相似度计算方法,基于信息量的计算方法,结合最低和最近公共祖先节点,提出了一种改进的相似度计算方法。通过相似度值与专家打分结果相比较的方法进行评价,实验结果表明本方法比传统方法更有效。Research on the functional similarity of genes or gene products is a focus of attention in Bioinformatics.Gene Ontology describes the attributes for gene and gene products.Go based term similarity calculation is of great benefit to analysis,comparison and prediction of gene function.This paper studies the semantic similarity calculation method based on gene ontology.Based on information content calculation method,combining the least common ancestor with most recent common ancestor nodes,the paper proposes an improved similarity calculation method.By comparing similarity values with expert score results to conduct the evaluation,experimental results show that this method is more effective than traditional methods.

关 键 词：基因本体 语义相似性 信息量 

分 类 号：TP391.4[自动化与计算机技术—计算机应用技术]