酶标法检测抗菌素残留量数据处理方法的探讨  

Discussion on Data Process of ELISA for Residual Antibiotics

作　　者：王静 潘海龙 袁良玉 严君喜 彭农 曲芙莲 

机构地区：[1]成都生物制品研究所质量检定室,成都610023

出　　处：《中国生物制品学杂志》2010年第9期1014-1017,共4页Chinese Journal of Biologicals

基　　金：国家科技支撑计划(计划编号:2008BAI54B07-02)

摘　　要：目的探讨酶标法检测抗菌素残留量的数据处理方法。方法采用卡那霉素和庆大霉素快速检测试剂盒(间接竞争ELISA法)检测乙型脑炎减毒活疫苗2种抗菌素的残留量;采用半对数直线、半对数二次曲线和对数直线3种方法,以最小二乘法对实验数据进行拟合处理,分析、比较实验点残差分布、残差平方和、相关系数和实验误差等指标。结果半对数直线拟合较差,半对数二次曲线和对数直线方程拟合较好;半对数二次曲线法中实验点在标准曲线两侧的分布较均匀,各点的残差平方和远小于半对数直线法;半对数二次曲线法和对数直线法所得标准曲线的相关系数较高,检测庆大霉素标准品的实验误差均小于半对数直线法。结论半对数二次曲线法和对数直线法均优于试剂盒推荐的半对数直线法,而在极限值处理、计算简便等方面对数直线法更优。Objective To discuss the data process of ELISA for residual antibiotics.Methods The residual kanamycin and gentamicin contents in live attenuated Japanese encephalitis vaccine were determined by rapid indirect competitive ELISA kits for the two kinds of antibiotics respectively.The obtained data were fitted by least square method according to semi-logarithmic linear,semilogarithmic quadratic curve and logarithmic linear equations respectively,based on which the distribution of test point residual errors,sum of square of residual errors,correlation coefficients and experimental errors were analyzed and compared.Results The data were unsuccessfully fitted according to semi-logarithmic linear equation,while successfully fitted according to semi-logarithmic quadratic curve and logarithmic linear equations.The test points were distributed evenly at two sides of standard semi-logarithmic quadratic curve,and the sum of square of test point residual errors was significantly smaller than those fitted according to semi-logarithmic linear equation.The correlation coefficients of standard curves obtained by semi-logarithmic quadratic curve and logarithmic linear regressions were higher,while the experimental errors in determination of gentamicin standard were smaller than that by semilogarithmic linear regression.Conclusion Both semi-logarithmic quadratic curve and logarithmic linear regressions were more suitable for data process of ELISA for residual antibiotics than semi-logarithmic linear regression recommended by the kit.However,logarithmic linear regression was more suitable for the process of limit value,by which the calculation was more simple.

关 键 词：酶标法 抗菌素 最小二乘法 半对数直线 半对数二次曲线 对数直线 

分 类 号：R195.1[医药卫生—卫生统计学]