Chemical Abstracts Service Chemical Registry System. 10. Registration of substances from pre-1965 indexes of Chemical Abstracts

The Chemical Abstracts Service Chemical Registry System, operating since 1965, uniquely identifies chemical substances on the basis of molecular structure. Chemical Abstracts Service is now registering chemical substances cited in indexes to Chemical Abstracts prior to 1965. This effort will result in several hundred thousand additional chemical structures, along with their names, being available for online searching in the Registry File. Both the newly registered substances and those already on file are being linked to their pre-1965 citations in Chemical Abstracts in a new file called CAOLD. In this effort the printed Formula Index entries are converted to computer-readable form by using optical character recognition with the data subsequently processed with existing computer programs.     

基于支持向量学习机方法的人体小肠吸收药物活性的预测

为了预测分子在人体小肠中的吸收,本文计算了表征分子的电子、拓扑、几何结构、分子形状等特征的102个分子描述符,用遗传算法变量选择方法使描述符减少到47个。体系共包含了230个化合物分子,69个不能被吸收(mA-),161个可以被吸收(HIA )。对建立的SVM模型,用5重交叉验证和独立测试集进行验证,预测正确率分别达到79．1％和77．1％,结果具有较好的一致性。在模型验证中,通过聚类分析方法组合训练集和测试集,保证了模型的稳定性,提高了建模效率。