Optimization of artificial neural network for retention modeling in high-performance liquid chromatography

An artificial neural network (ANN) model for the prediction of retention times in high-performance liquid chromatography (HPLC) was developed and optimized. A three-layer feed-forward ANN has been used to model retention behavior of nine phenols as a function of mobile phase composition (methanol-acetic acid mobile phase). The number of hidden layer nodes, number of iteration steps and the number of experimental data points used for training set were optimized. By using a relatively small amount of experimental data (25 experimental data points in the training set), a very accurate prediction of the retention (percentage normalized differences between the predicted and the experimental data less than 0.6%) was obtained. It was shown that the prediction ability of ANN model linearly decreased with the reduction of number of experiments for the training data set. The results obtained demonstrate that ANN offers a straightforward way for retention modeling in isocratic HPLC separation of a complex mixture of compounds widely different in pK_a and log K_ow values.     

基于支持向量机的药物诱导磷脂质病预测模型

本文应用一种组合遗传算法和共轭梯度法的支持向量机(GA-CG-SVM)方法建立了药物诱导磷脂质病分类预测模型.首先对描述符进行了优化,选出了19个描述符用于模型的构建,所建模型对训练集的预测准确率为81.6%,对测试集的预测精度为87.5%,说明所建SVM分类模型不仅能正确预测训练集药物诱导的磷脂质病,也对其他化合物具...     

A neural network based virtual high throughput screening test for the prediction of CNS activity

A virtual high throughput screening test to identify potentially CNS-active drugs has been developed. Discrimination was based on the knowledge available in databases containing CNS-active (Cipsline from Prous Science) and inactive compounds (Chemical Directory from Sigma-Aldrich). Molecular structures were represented using 2D Unit y fingerprints and a feedforward neural network was trained to classify molecules regarding their CNS activity. The parameterized network was validated by reclassification of the training set elements, by the classification of a test set preselected from the Prous database, and also by the prediction of activity for known CNS drugs not used in the training set but available in the Medchem database (Daylight). These tests revealed that our neural net recognized at least 89% of CNS-active compounds and would be suitable for use in our virtual screening protocol.     

Optimization of artificial neural networks used for retention modelling in ion chromatography

The aim of this work is the development of an artificial neural network model, which can be generalized and used in a variety of applications for retention modelling in ion chromatography. Influences of eluent flow-rate and concentration of eluent anion (OH-) on separation of seven inorganic anions (fluoride, chloride, nitrite, sulfate, bromide, nitrate, and phosphate) were investigated. Parallel prediction of retention times of seven inorganic anions by using one artificial neural network was applied. MATLAB Neural Networks ToolBox was not adequate for application to retention modelling in this particular case. Therefore the authors adopted it for retention modelling by programming in MATLAB metalanguage. The following routines were written; the division of experimental data set on training and test set; selection of data for training and test set; Dixon's outlier test; retraining procedure routine; calculations of relative error. A three-layer feed forward neural network trained with a Levenberg-Marquardt batch error back propagation algorithm has been used to model ion chromatographic retention mechanisms. The advantage of applied batch training methodology is the significant increase in speed of calculation of algorithms in comparison with delta rule training methodology. The technique of experimental data selection for training set was used allowing improvement of artificial neural network prediction power. Experimental design space was divided into 8-32 subspaces depending on number of experimental data points used for training set. The number of hidden layer nodes, the number of iteration steps and the number of experimental data points used for training set were optimized. This study presents the very fast (300 iteration steps) and very accurate (relative error of 0.88%) retention model, obtained by using a small amount of experimental data (16 experimental data points in training set). This indicates that the method of choice for retention modelling in ion chromatography is the artificial neural network.     

Probabilistic assessment of biodegradability based on metabolic pathways: CATABOL System

A novel mechanistic modeling approach has been developed that assesses chemical biodegradability in a quantitative manner. It is an expert system predicting biotransformation pathway working together with a probabilistic model that calculates probabilities of the individual transformations. The expert system contains a library of hierarchically ordered individual transformations and matching substructure engine. The hierarchy in the expert system was set according to the descending order of the individual transformation probabilities. The integrated principal catabolic steps are derived from set of metabolic pathways predicted for each chemical from the training set and encompass more than one real biodegradation step to improve the speed of predictions. In the current work, we modeled O 2 yield during OECD 302 C (MITI I) test. MITI-I database of 532 chemicals was used as a training set. To make biodegradability predictions, the model only needs structure of a chemical. The output is given as percentage of theoretical biological oxygen demand (BOD). The model allows for identifying potentially persistent catabolic intermediates and their molar amounts. The data in the training set agreed well with the calculated BODs ( r 2 =0.90) in the entire range i.e. a good fit was observed for readily, intermediate and difficult to degrade chemicals. After introducing 60% ThOD as a cut off value the model predicted correctly 98% ready biodegradable structures and 96% not ready biodegradable structures. Crossvalidation by four times leaving 25% of data resulted in Q 2 =0.88 between observed and predicted values. Presented approach and obtained results were used to develop computer software for biodegradability prediction CATABOL.     

Use of PLS Discriminant Analysis for Revealing the Absence of a Compound in an Electron Ionization Mass Spectral Database

A mathematical model is proposed for revealing the absence of a compound to be identified in an electron impact mass spectral library. The mathematical model (developed based on PLS Discriminant Analysis) can be represented as a “black box” which provides an answer whether a compound to be sought is absent or present in a database. The match factors of top ten candidates among the possible ones were used as input data. More than 5000 objects (mass spectra) were used at the steps of training, validation, and testing. The developed classification model provides correct prediction (of whether a compound is absent from the library) in 28.4% cases, while only 1.2% of compounds present in the database were incorrectly classified as the absent ones.     

Comparative study to predict toxic modes of action of phenols from molecular structures

Quantitative structure–activity relationship models for the prediction of mode of toxic action (MOA) of 221 phenols to the ciliated protozoan Tetrahymena pyriformis using atom-based quadratic indices are reported. The phenols represent a variety of MOAs including polar narcotics, weak acid respiratory uncouplers, pro-electrophiles and soft electrophiles. Linear discriminant analysis (LDA), and four machine learning techniques (ML), namely k-nearest neighbours (k-NN), support vector machine (SVM), classification trees (CTs) and artificial neural networks (ANNs), have been used to develop several models with higher accuracies and predictive capabilities for distinguishing between four MOAs. Most of them showed global accuracy of over 90%, and false alarm rate values were below 2.9% for the training set. Cross-validation, complementary subsets and external test set were performed, with good behaviour in all cases. Our models compare favourably with other previously published models, and in general the models obtained with ML techniques show better results than those developed with linear techniques. We developed unsupervised and supervised consensus, and these results were better than our ML models, the results of rule-based approach and other ensemble models previously published. This investigation highlights the merits of ML-based techniques as an alternative to other more traditional methods for modelling MOA.