Analysis of the Internal Representations Developed by Neural Networks for Structures Applied to Quantitative Structure-Activity Relationship Studies of Benzodiazepines

An application of recursive cascade correlation (CC) neural networks to quantitative structure-activity relationship (QSAR) studies is presented, with emphasis on the study of the internal representations developed by the neural networks. Recursive CC is a neural network model recently proposed for the processing of structured data. It allows the direct handling of chemical compounds as labeled ordered directed graphs, and constitutes a novel approach to QSAR. The adopted representation of molecular structure captures, in a quite general and flexible way, significant topological aspects and chemical functionalities for each specific class of molecules showing a particular chemical reactivity or biological activity. A class of 1,4-benzodiazepin-2-ones is analyzed by the proposed approach. It compares favorably versus the traditional QSAR treatment based on equations. To show the ability of the model in capturing most of the structural features that account for the biological activity, the internal representations developed by the networks are analyzed by principal component analysis. This analysis shows that the networks are able to discover relevant structural features just on the basis of the association between the molecular morphology and the target property (affinity).

[1]  D. Richman,et al.  Inhibition of HIV replication in acute and chronic infections in vitro by a Tat antagonist. , 1991, Science.

[2]  Scott E. Fahlman,et al.  The Recurrent Cascade-Correlation Architecture , 1990, NIPS.

[3]  Alessandro Sperduti,et al.  Extended Cascade-Correlation for Syntactic and Structural Pattern Recognition , 1996, SSPR.

[4]  Alessio Micheli,et al.  Quantitative structure-activity relationships of Benzodiazepines by recursive cascade correlation , 1998, 1998 IEEE International Joint Conference on Neural Networks Proceedings. IEEE World Congress on Computational Intelligence (Cat. No.98CH36227).

[5]  Johann Gasteiger,et al.  Neural Networks for Chemists: An Introduction , 1993 .

[6]  George M. Whitesides,et al.  FEED-FORWARD NEURAL NETWORKS IN CHEMISTRY : MATHEMATICAL SYSTEMS FOR CLASSIFICATION AND PATTERN RECOGNITION , 1993 .

[7]  C. Hansch,et al.  Quantitative Structure‐Activity Relationships of the Benzodiazepines. A Review and Reevaluation. , 1995 .

[8]  L. Sternbach,et al.  The benzodiazepine story. , 1979, Journal of medicinal chemistry.

[9]  Jiri Pospichal,et al.  Application of neural networks in chemistry. Prediction of product distribution of nitration in a series of monosubstituted benzenes , 1991 .

[10]  P Jenner,et al.  The Benzodiazepines: from Molecular Biology to Clinical Practice , 1985 .

[11]  Ruisheng Zhang,et al.  Neural Network-Topological Indices Approach to the Prediction of Properties of Alkene , 1997, J. Chem. Inf. Comput. Sci..

[12]  Erik De Clercq,et al.  Potent and selective inhibition of HIV-1 replication in vitro by a novel series of TIBO derivatives , 1990, Nature.

[13]  C. Hansch,et al.  p-σ-π Analysis. A Method for the Correlation of Biological Activity and Chemical Structure , 1964 .

[14]  Christian Lebiere,et al.  The Cascade-Correlation Learning Architecture , 1989, NIPS.

[15]  C. Hansch,et al.  QUANTITATIVE STRUCTURE-ACTIVITY RELATIONSHIPS OF THE BENZODIAZEPINES. A REVIEW AND REEVALUATION , 1994 .

[16]  H. Sugano,et al.  Synthesis and central nervous system actions of thyrotropin-releasing hormone analogues containing a dihydroorotic acid moiety. , 1990, Journal of medicinal chemistry.

[17]  Gérard Dreyfus,et al.  Toward a Principled Methodology for Neural Network Design and Performance Evaluation in QSAR. Application to the Prediction of LogP , 1998, J. Chem. Inf. Comput. Sci..

[18]  I. Jolliffe.  Principal Component Analysis , 2002 .

[19]  J Desmyter,et al.  Potent and selective inhibition of HIV-1 replication in vitro by a novel series of tetrahydro-imidazo[4,5,1-JK][1,4]-benzzodiazepin-2(1H)-one and -thione (TIBO) derivatives , 1990 .

[20]  B. E. Evans,et al.  Benzodiazepine gastrin and brain cholecystokinin receptor ligands: L-365,260. , 1989, Journal of medicinal chemistry.

[21]  James Devillers,et al.  Neural Networks in QSAR and Drug Design , 1996 .

[22]  E. Costa,et al.  The Benzodiazepines: From molecular biology to clinical practice , 1983 .

[23]  S. Free,et al.  A MATHEMATICAL CONTRIBUTION TO STRUCTURE-ACTIVITY STUDIES. , 1964, Journal of medicinal chemistry.

[24]  Ajay.  A unified framework for using neural networks to build QSARs. , 1993, Journal of medicinal chemistry.

[25]  D. Villemin,et al.  Use of a neural network to determine the boiling point of alkanes , 1994 .

[26]  Alessandro Sperduti,et al.  Supervised neural networks for the classification of structures , 1997, IEEE Trans. Neural Networks.

[27]  H Ichikawa,et al.  Neural networks applied to quantitative structure-activity relationship analysis. , 1990, Journal of medicinal chemistry.

[28]  R. Lenox,et al.  Platelet-activating factor-induced aggregation of human platelets specifically inhibited by triazolobenzodiazepines. , 1984, Science.

[29]  Keith L. Peterson.  Quantitative Structure-Activity Relationships in Carboquinones and Benzodiazepines Using Counter-Propagation Neural Networks , 1995, J. Chem. Inf. Comput. Sci..

[30]  Gerald M. Maggiora,et al.  Applications of neural networks in chemistry. 1. Prediction of electrophilic aromatic substitution reactions , 1990, J. Chem. Inf. Comput. Sci..

[31]  R. M. Muir,et al.  Correlation of Biological Activity of Phenoxyacetic Acids with Hammett Substituent Constants and Partition Coefficients , 1962, Nature.