一种计算水溶解度的经验加合模型的适用范围与局限

我们发展了一种用于预测有机小分子化合物水溶解度(logS)的经验方法XLOGS. 它本质上是一种加合模型, 采用83种原子/基团类型和3个校正因子作为模型的描述符.该方法还可以根据一个合适的参照分子的logS实验值来计算未知化合物的logS值. 我们将XLOGS模型在由4171个化合物组成的训练集上进行了参数化, 多元线性回归获得的相关系数R²和标准偏差SD分别为0.82和0.96单位. 将该训练集进一步分为仅含液体化合物和仅含固体化合物的两个子集. XLOGS模型在这两个子集上的回归结果显示前者优于后者(标准偏差分别为0.65单位和0.94单位). 还利用log1/S和logP(脂水分配系数)之间的差值来研究XLOGS方法在液体和固体化合物数据集上的表现. 研究结果表明: XLOGS等加合法模型更适合应用于这两者差值接近于0的化合物. 我们还将XLOGS和其他三种流行的logS计算模型(包括Qikprop, MOE-logS和ALOGPS)在一个含有132个类药化合物的独立测试集上进行了比较. 总体而言, 我们的研究结果为加合法模型在水溶解度预测方面的合理应用提供了指导.     

计算有机化合物脂水分配系数的新方法

A new method is presented for the calculation of octanol/water partition coefficients. On the basis of summation of atomic contributions, our algorithm, namely; XLOGP, also incorporate correction factors into the calculation. Multivariate regression analysis was performed on a training database of 1831 organic compounds with diverse structures to give the final model. The correlation coefficients for the whole set fitting is 0.968 and the standard deviation is 0.37. The result shows that our model is accurate enough for logP estimation in QSAR studies. Compared to other similar approaches, our method gives better results and is more convenient to use.     

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。     

Indirect similarity based methods for effective scaffold-hopping in chemical compounds

Methods that can screen large databases to retrieve a structurally diverse set of compounds with desirable bioactivity properties are critical in the drug discovery and development process. This paper presents a set of such methods that are designed to find compounds that are structurally different to a certain query compound while retaining its bioactivity properties (scaffold hops). These methods utilize various indirect ways of measuring the similarity between the query and a compound that take into account additional information beyond their structure-based similarities. The set of techniques that are presented capture these indirect similarities using approaches based on analyzing the similarity network formed by the query and the database compounds. Experimental evaluation shows that most of these methods substantially outperform previously developed approaches both in terms of their ability to identify structurally diverse active compounds as well as active compounds in general.