Neural Network Correlations

Buenz, A.P., Braun, B. and Janowsky, R. 1999, Quantitative structure-property relationships and neural networks Correlation and prediction of physical properties of pure components and mixtures from molecular structure. Fluid Phase Equilibria 158-160,367. 

The new neural network correlations give mean absolute errors of 1.06 for cetane number and 0.003 for density, which represent significant improvements over the correlations developed in the first attempt. Work on improving these correlations and developing correlations for the other properties available in our database continues. 

Chirality codes are used to represent molecular chirality by a fixed number of de-.scriptors. Thc.se descriptors can then be correlated with molecular properties by way of statistical methods or artificial neural networks, for example. The importance of using descriptors that take different values for opposite enantiomers resides in the fact that observable properties are often different for opposite enantiomers. 

The possibilities for the application for neural networks in chemistry arc huge [10. They can be used for various tasks for the classification of structures or reactions, for establishing spcctra-strncturc correlations, for modeling and predicting biological activities, or to map the electrostatic potential on molecular surfaces. 

Association deals with the extraction of relationships among members of a data set. The methods applied for association range from rather simple ones, e.g., correlation analysis, to more sophisticated methods like counter-propagation or back-propagation neural networks (see Sections 9.5.5 and 9.5.7). 

Neural networks have been applied to IR spectrum interpreting systems in many variations and applications. Anand [108] introduced a neural network approach to analyze the presence of amino acids in protein molecules with a reliability of nearly 90%. Robb and Munk [109] used a linear neural network model for interpreting IR spectra for routine analysis purposes, with a similar performance. Ehrentreich et al. [110] used a counterpropagation network based on a strategy of Novic and Zupan [111] to model the correlation of structures and IR spectra. Penchev and co-workers [112] compared three types of spectral features derived from IR peak tables for their ability to be used in automatic classification of IR spectra. 

Several methods have been developed for establishing correlations between IR vibrational bands and substructure fragments. Counterpropagation neural networks were used to make predictions of the full spectra from RDF codes of the molecules. 

In this approach, connectivity indices were used as the principle descriptor of the topology of the repeat unit of a polymer. The connectivity indices of various polymers were first correlated directly with the experimental data for six different physical properties. The six properties were Van der Waals volume (Vw), molar volume (V), heat capacity (Cp), solubility parameter (5), glass transition temperature Tfj, and cohesive energies ( coh) for the 45 different polymers. Available data were used to establish the dependence of these properties on the topological indices. All the experimental data for these properties were trained simultaneously in the proposed neural network model in order to develop an overall cause-effect relationship for all six properties. 

Neural networks are extensively used to develop nonparametric models and are now the method of choice when electronic noses are used to analyze complex mixtures, such as wines and oils.5 Judgments made by the neural network cannot rely on a parametric model that the user has supplied because no model is available that correlates chemical composition of a wine to the wine s taste. Fortunately, the network can build its own model from scratch, and such models often outperform humans in determining the composition of oils, perfumes, and wines. 

The role of an artificial neural network is to discover the relationships that link patterns of input data to associated output data. Suppose that a database contains information on the structure of many potential drug molecules (the input) and their effectiveness in treating some specific disease (the output). Since the clinical value of a drug must in some way be related to its molecular structure, correlations certainly exist between structure and effectiveness, but those relationships may be very subtle and deeply buried. 

Another way for BOD estimation is the use of sensor arrays [37]. An electronic nose incorporating a non-specific sensor array of 12 conducting polymers was evaluated for its ability to monitor wastewater samples. A statistical approach (canonical correlation analysis) showed a linear relationship between the sensor responses and BOD over 5 months for some subsets of samples, leading to the prediction of BOD values from electronic nose analysis using neural network analysis. 

A classical Hansch approach and an artificial neural networks approach were applied to a training set of 32 substituted phenylpiperazines characterized by their affinity for the 5-HTiA-R and the generic arAR [91]. The study was aimed at evaluating the structural requirements for the 5-HTiA/ai selectivity. Each chemical structure was described by six physicochemical parameters and three indicator variables. As electronic descriptors, the field and resonance constants of Swain and Lupton were used. Furthermore, the vdW volumes were employed as steric parameters. The hydrophobic effects exerted by the ortho- and meta-substituents were measured by using the Hansch 7t-ortho and n-meta constants [91]. The resulting models provided a significant correlation of electronic, steric and hydro-phobic parameters with the biological affinities. Moreover, it was inferred that the... 

In addition to methods that rely on experimental inputs, there are prediction tools that are based on structure alone. A number of different techniques have been used to correlate a variety of structure-derived descriptors with observed solubilities, including linear regression [20-25] and neural networks [26-36]. Interestingly, despite... 

---