论文部分内容阅读
以简单配比系数为相似性指标对１５种发育毒物体外筛选试验进行指标型系统聚类分析。结果提示最长、最短和平均距离法３种聚类方法中以最长距离法的聚类效果最为理想，符合发育毒物毒作用的细胞学机制。最终聚类结果为５大类，其中以Ａ／Ｄ比值和以细胞生长为指标的方法聚为一类；以细胞聚集、细胞生长和细胞分化３个终点为联合指标的鸡胚神经视网膜细胞试验和啮齿类动物的全胚胎培养归为一类。但以反映细胞分化和细胞间相互作用以及胚胎形成抑制为终点的方法并未完全归为一类，说明这些单一的终点不可能全面反映细胞增殖、分化和相互作用等相互依赖的细胞行为，因而提示应用其中的两个终点进行组合并不能提高预测价值。为了全面提高发育毒物体外筛选试验的预测价值，应结合各种方法的特异度和灵敏度，并根据经济有效和合理的原则，选择其中的３～５种进行组合
Index system was used to analyze the in vitro screening test of 15 kinds of developmental toxicants with the simple matching coefficient as the similarity index. The results suggest that the longest, shortest and average distance method clustering method among the three kinds of clustering method with the longest distance method is the most ideal, in line with the development of toxicological cytotoxic mechanisms. The final clustering results were divided into five groups, of which the A / D ratio and cell growth were clustered into one category. The chick embryo neural retinal cells assay with three endpoints of cell aggregation, cell growth and cell differentiation And rodent whole embryo culture fall into one category. However, the methods that reflect the endpoints of cell differentiation and cell-cell interactions and inhibition of embryogenesis are not completely classified, suggesting that these single endpoints can not fully reflect the interdependent cellular behaviors such as cell proliferation, differentiation and interaction and thus Suggesting that combining two of these end points does not improve the predictive value. In order to comprehensively improve the predictive value of developmental in vitro screening tests, we should combine the specificity and sensitivity of various methods and select 3 to 5 of them according to the principle of economic efficiency and reasonableness