L. Buydens

An expert system for the selection of criteria for selectivity optimization in high-pressure liquid chromatography

SummarySystematic procedures for the optimization of chromatographic selectivity require objective criteria to characterize the quality of separation in a chromatogram. Numerous criteria have been suggested. Different criteria yield different results and the choice will depend on a large number of factors. It is genuinely difficult to select the most suitable criterion in a particular situation. For these reasons, an expert system has been developed to assist chromatographers in the selection of optimization criteria. A structured representation of the required knowledge and its implementation in an expert-system shell are presented in this paper.

Expert system for the selection of high-performance liquid chromatographic methods in pharmaceutical analysis : Validation of the rules for the selection of the mobile phase

The rules for the selection of the mobile phase and the validation performed on 44 pharmaceutical preparations, containing one to five active compounds, are described. These rules are incorporated into an expert system, called LABEL, for the selection of high-performance liquid chromatographic methods in pharmaceutical analysis. A single stationary phase type is used, namely a nitrile or cyanopropyl (CN) column, which can be used in both normal-phase (NP) and reversed-phase (RP) chromatography. Three mobile phase systems were evaluated on this column type: NP, RP with water and RP with buffer. LABEL selects one of these three systems on the basis of the rules incorporated for the mobile phase selection, checks if the addition of ion-suppressing agents to the eluting agent is necessary and finally gives the starting composition of the mobile phase in each of the three systems. For this selection the number of compounds in the sample, the acid-base properties and the hydrophobicity of the solutes are the more important factors. The validation of the rules on 44 pharmaceutical preparations resulted in an immediate success in 82% of all cases. In half of the remaining cases, the system proposed can be adapted with a minor change in conditions, so that it can also be used in practice.

The use of principal components analysis for the investigation of an organic air pollutants data set

Abstract Principal components analysis (PCA) was used for the investigation of an air pollutants data base. The data set consists of nearly 400 measurements of 26 gaseous organic compounds and meteorological data. The measurements were carried out at four different places in The Netherlands. PCA is a simple way to display visually most of the total variation in a few dimensions. It is helpful in the identification of outliers, the recognition of sources and the investigation of meteorological effects.

Expert system for the selection of high-performance liquid chromatographic methods for the analysis of drugs

Abstract An expert system for the selection of high-performance liquid chromatographic methods is described for the label claim analysis of drugs in pharmaceutical formulations. The system contains knowledge for the selection of a suitable detection mode (UV detection or electrochemical detection in the oxidative nmode), an appropriate chromatographic mode (reversed-phase with water, reversed-phase with buffers or normal phase) and the starting mobile phase compositions in each chromatographic mode. The chromatography is performed on a single type of column, namely a cyanopropyl column. The implementation of the knowledge in the commercially available expert system shell (KES) is also described. As a knowledge representation method, production rules are used.

Evaluation of the expert system shells EX-TRAN and TIMM as rule-building tools for classification purposes

Abstract This article describes results obtained with two rule-building expert systems using simulated data sets. It was found that certain data structures (e.g. strongly correlated data or data containing outliers) cannot be handled in an optimal way. In some cases (e.g. with strongly overlapping data) one can observe from the rule set that the classification is not robust. This requires expertise however. Our results lead us to conclude that these expert systems do not yet possess sufficient expertise to be used without problems by non-experts or in conjunction with analytical instruments.

Chemometrics in pharmaceutical analysis

Chemometrics is a science where chemistry and pharmaceutical science meet statistics and software. The primary focus of chemometrics involves the use of mathematical or software procedures in particular, both to develop analytical methods and to analyse the signals and results obtained. This paper focusses on chromatography and on how chemometrics has been applied to chromatographic problems in pharmaceutical and biomedical analysis. Examples of several chemometric techniques are given and recent developments in the use of optimization methods, regression methodology, multivariate analysis and expert systems are discussed.

Feasibility study concerning the use of expert systems for the development of procedures in pharmaceutical analysis

The feasibility of using expert systems for the development of analytical procedures is investigated. A system for the computer generation of procedures to determine active drug substances in commercial formulations is proposed. It is shown that in nearly 85% of the cases investigated the present system immediately yields a correct procedure or conclusion. It is concluded that selecting methods and developing procedures with the use of expert systems is difficult but feasible.

Knowledge representation for the selection of methods in analytical chemistry

Abstract Expert systems are gaining rapidly in importance in several areas of science, and in chemistry in particular. One of the most important factors that determines the success of an expert system is the representation of the knowledge. To understand better the suitability of the different knowledge representation techniques for the selection of methods in analytical chemistry, a test knowledge base was developed for the high-performance liquid chromatographic analysis of pharmaceutical compounds. This knowledge base has been used to study which features of knowledge representation are necessary to describe analytical chemical knowledge in a natural and efficient way. It is concluded that no single representation method is optimal. A combination of production rules and frame structures seems to be much more suitable. The possibility of consulting external databases and programs is also a very important aspect.

An expert system for the development of analytical procedures: UV spectrophotometric determination of pharmaceutically active substances in tablets

Abstract Buydens, L., Detaevernier, M., Tombeur, D. and Massart, D.L., 1986. An expert system for the development of analytical procedures: UV spectrophotometric determination of pharmaceutically active substances in tablets. Chemometrics and Intelligent Laboratory Systems , 1: 99–108. An expert system developed using KES is presented. It is explained how such an expert system was developed and validation results are presented.

Gas chromatographic behaviour and pharmacological activity of neuroleptica

Biological activity in vitro, quantified as equilibrium inhibition constants to the dopamine receptor, of a series of neuroleptica are correlated with parameters describing polar and apolar interactions of these molecules with the receptor. Gas-chromatographic retention indices on stationary phases of different polarity are compared with parameters that are classically used in such quantitative structure/activity studies. To describe a polar interactions, a series of classical parameters such as several valence molecular indices and log k′ in a reversed-phase h.p.l.c. system are included; retention indices on the apolar stationary phase, OV101, are used as the gas-chromatographic (g.c.) parameter. To take the more specific polar interactions into account, quantum-chemical charge parameters such as the dipole moment and the charge on atoms directly involved in the interaction were calculated. Retention indices on the more polar phase OV17 are taken as the g.c. parameter for polar interactions. It is shown that the retention indices on OV101 can replace classical parameters describing aspecific or apolar interactions. The retention indices of OV17 do not correlate with biological activity and are worse than the charge parameters.