一种加入相反表型的语义相似度算法用于罕见病辅助诊断  

A Semantic Similarity Algorithm Using Opposite Phenotypes for the Auxiliary Diagnosis of Rare Diseases

作　　者：王培萱 吕晖 Wang Peixuan;Lu Hui(School of Life Science and Biotechnology,Shanghai Jiao Tong University,Shanghai,200240)

机构地区：[1]上海交通大学生命科学技术学院,上海200240

出　　处：《基因组学与应用生物学》2020年第12期5868-5873,共6页Genomics and Applied Biology

基　　金：国家卫生计生委医药卫生科技发展研究中心“精准医学研究”重点专项项目(2018YFC0910500)资助。

摘　　要：在日常的临床诊断中,医生常常根据入院病人的异常表型症状进行诊断,这种方式需要医生具备足够的经验才能完成。人工智能的出现使临床诊断出现的新的思路,只要获得足够的训练样本及正确的知识库,就可以帮助医生进行辅助诊断。人类表型本体可以将临床数据转化为结构化的表示形式,并以系统的方式组织管理这些数据。通过使用这些结构化的表型数据,我们可以使用不同的算法辅助诊断罕见疾病和先天性遗传疾病。基于表型的罕见病辅助诊断需要根据表型数据的语义相似度进行打分,将表型集合和疾病的相似度计算出来,将疾病根据相似度得分进行排名,从而得到预测结果的排名。本研究使用公共开源数据库HPO中的表型数据,通过使用添加相反表型的信息,提出了一个新的语义相似度算法,该算法将表型与疾病进行了信息量的量化。本研究使用了模拟数据,与其他算法进行了比较,验证了该算法预测效果。结果显示添加相反表型的算法在模拟数据集上显示了较好的效果,该算法具有合理性和优越性。In routine clinical diagnosis,doctors often diagnose disease based on abnormal phenotypic symptoms of patients,which requires sufficient experience from the doctors.The emergence of artificial intelligence provides new ideas for clinical diagnosis.As long as sufficient training samples and the correct knowledge base are obtained,computers can help doctors to assist diagnosis.Human Phenotype Ontology can transform clinical data into structured representations and organize these data in a systematic way.By using these structured phenotypic data,people can use different algorithms to aid in the diagnosis of rare diseases and congenital genetic diseases.The way of phenotypic diagnosis is to calculate the semantic similarity of phenotypic data and then rank the scores.Disease is ranked according to the similarity score,thereby the predicted results can be obtained.In this paper,the public open source database HPO is used.A new semantic similarity algorithm is proposed by using the information of adding the opposite phenotype.The method quantifies the information of phenotypes and diseases.The simulation data is used and compared with other algorithms to verify the prediction effect of the algorithm.The results show that the algorithm of adding the opposite phenotype shows good results on the simulated data set.And the algorithm is rational and superior.

关 键 词：相反表型 语义相似度 辅助诊断 

分 类 号：TP391.1[自动化与计算机技术—计算机应用技术] R44[自动化与计算机技术—计算机科学与技术]