QSPR-Polymer

D. Case IV Integrating the Concepts of Graphical Theory and ANNs for Polymer Property Predictions in QSPR... 

Keywords Information systems Machine learning Ontology Polymer markup language Polymer informatics QSPR RDF Semantic web... 

A more common use of informatics for data analysis is the development of (quantitative) structure-property relationships (QSPR) for the prediction of materials properties and thus ultimately the design of polymers. Quantitative structure-property relationships are multivariate statistical correlations between the property of a polymer and a number of variables, which are either physical properties themselves or descriptors, which hold information about a polymer in a more abstract way. The simplest QSPR models are usually linear regression-type models but complex neural networks and numerous other machine-learning techniques have also been used. 

Two very simple types of QSPR have been developed early on in the evolution of polymer property prediction, namely van Krevelen s group contribution methods [122] and Bicerano s system [123], which mainly relies on the use of topological descriptors. Group contributions regard the overall properties of the polymer as the scalar sum of the properties of the chemical groups contained in the molecules making up the polymer. 

While both the Bicerano and van Krevelen systems model a significant number of polymer properties, most QSPR studies have focused on only a small number of key properties (which is mainly correlated to the availability of data for model development). 

Liu and Zhong introduced a number of QSPR models based on molecular connectivity indices [151, 152], In a first iteration, the researchers developed polymer-dependent correlations descriptors were calculated for a set of solvents and models were developed per polymer type [151], Polymer classes under consideration were polystyrene, polyethylene, poly-1-butene, poly-l-pentene, poly(4-methyl-l-pentene), polydimethylsiloxane, and polyisobutylene. As the authors fail to provide any validation for their models, it is difficult to asses their predictive power. In a subsequent iteration and general expansion of this study, mixed and therefore more general models based on the calculated connectivity indices of both solvent and polymers were developed. While it is unclear from the paper which polymer representation was used for the calculation of the connectivity indices, the best regression model (eight parameter model) yields only acceptable predictive power (R = 0.77, = 0.77, s = 34.47 for the training set, R = 0.75... 

There have been isolated QSPR studies of a number of other polymer properties. These include the dielectric constant [144], the dielectric dissipation factor (tan 8) [168], the solubility parameter [169], the molar thermal decomposition function [170], the vitrification temperature of polyarylene oxides [171], and quantities relating to molecularly imprinted polymers [172, 173]. The interested reader is referred to the literature for further information. 

Katritzky AR, SUd S, Lobanov V et al. (1998) Quantitative structure-property relationship (QSPR) correlation of glass transition temperatures of high molecular weight polymers. J Chem Inf Comput Sci 38 300-304... 

Xu J, Chen B, Zhang Q et al. (2004) Prediction of refractive indices of linear polymers by a four descriptor QSPR model. Polymer 45 8651-8659... 

Xu J, Liang H, Chen B et al. (2008) Linear and nonlinear QSPR models to predict refractive indices of polymers from cyclic dimer structures. Chemom Intell Lab Syst 92 152-156... 

Melagraki G, Afantitis A, Sarimveis H et al. (2007) A novel QSPR model for predicting theta (lower critical solution temperature) in polymer solutions using molecular descriptors. J Mol Model 15 55-64... 

Afantitis A, Melagraki G, Satimveis H et al. (2006) Prediction of intrinsic viscosity in polymer-solvent combinations using a QSPR model. Polymer 47 3240-3248... 

Methods based on quantitative structure-property relationships (QSPR) have been available for some time now and have become more or less standard empirical techniques since the appearance in the literature of van Krevelen s now classic book currently in its third edition. All these methodologies take advantage of the vast databases of experimental data that have been accumulated over the years by mainly industrial but also by academic laboratories. The methodology described by van Krevelen is based on group contribution methods and it works satisfactorily for those polymers for which information on group contributions exists. 

In some cases selected atoms in molecules could be marked with special labels, indicating their particular role in a modeled property. Some examples are (i) local properties, such as atomic charges or NMR chemical shifts, which should always be attributed to a given atom(s), (ii) anchor atoms in the given scaffold to which substituents are attached (Figure 1.13), (iii) atoms forming a main chain in polymers and (iv) reaction centers in a set of reactions. Zefirov et al. have applied labeling in QSPR studies of chemical NMR shifts and... 

Katritzky, A.R., Sild, S. and Karelson, M. (1998a). Correlation and Prediction of the Refractive Indices of Polymers by QSPR. J.ChenuInf.Comput.ScL, 38,1171-1176. 

QSAR Quantitative structure-activity relationships (term used for ordinary molecules). QSPR Quantitative structure-property relationships (term used for polymers). 

Additive (group contribution) methods have a long tradition of successful use in predicting the properties of both ordinary molecules and macromolecules (polymers). They have formed the backbone of the quantitative structure-activity relationships (QSAR) [1,2] used to predict the chemical reactivity and the biological activity of molecules in medicinal and agricultural chemistry. They have also been used extensively in many quantitative structure-property relationships (QSPR) developed for the physical and chemical properties of polymers. 

Van Krevelen [3] published the classic textbook on group contribution techniques in the QSPR of polymers. This book contains a compendium of useful QSPR relationships for polymers, as well as tables of large amounts of experimental data. The information contained in this book has been extremely valuable in the development of many of our correlations. We will, consequently, refer very frequently to this book as well as to a later review article by van Krevelen [4] containing a significant amount of additional or revised information. While we will refer to the third edition of van Krevelen s book (published in 1990), the reader should be aware that some of our correlations developed with his hook as a key resource used its second edition (published in 1976). In general, we did not revise correlations that we developed by using the information in the second edition as a resource, unless the third edition contained information suggesting qualitatively different conclusions. 

---