基于多波长透射光谱对多种细菌的鉴别方法研究

Study on the identification method of multiple bacteria based on multi-wavelength transmission

目的基于紫外分光光度计能利用多波长透射光谱有效反应细菌的内部结构以及化学成分,对快速检测水体细菌,保护人类的生命健康具有重要意义。方法利用S1(相关系数法)、S2(相似学原理)、S3(皮尔森相关系数)、S4(余弦相似度)、S5(欧几里得距离表示的相似度)、S6(曼哈顿距离表示的相似度)6种相似度算法以及max-min归一化数据预处理方法对芦荟微球菌(Micrococcus aloe)、金黄色葡萄球菌(Staphylococcus aureus)、球形赖氨酸芽胞杆菌(Bacillus lysine globulus)、苏云金芽胞杆菌(Bacillus thuringiensis)、水莱茵默菌(Rheinheimera aquimaris)等5种细菌的数据进行研究;并利用S2、S5、S63种相似度较好的计算方法研究了随机样本与它相应的模板的识别率。结果S2、S5、S6、拥有2种最大值指向、3种最大值指向5种方法对5种细菌的总体识别率分别为90%、100%、90%、92%和100%。结论本文微生物的检测方法具有一定的可靠性。

Objective Based on ultraviolet spectrophotometer can use multi-wavelength transmission spectrum to effectively reflect the internal structure and chemical composition of bacteria,which is of great significance for rapid detection of bacteria in water and protection of human life and health.Methods S1(correlation coefficient method),S2(similar principles),S3(Pearson correlation coefficient),S4(cosine similarity),S5(Euclidean distance similarity),S6(Manhattan distance similarity),the 6 kinds of similarity algorithm,and the Max-min normalized data preprocessing method of aloe Micrococcus(Micrococcus aloe),Staphylococcus aureus,Bacillus lysine globulus,Bacillus thuringiensisandRheinheimera aquimariswere used to study data.The recognition rate between the random sample and its corresponding template is studied by using the calculation methods of S2,S5 and S6 with better similarity.Results The overall recognition rates of S2,S5,S6 and S5 were 90%,100%,90%,92%and 100%,respectively.Conclusion The method of microorganism detection in this paper is reliable.