Expert model for evaluating the acute toxicity of structurally related compounds - azatricyclononane derivatives
"27 Central research Institute" of the Ministry of defense of the Russian Federation
Brief summary
In this paper, the possibility of predicting the toxic properties of azabrendan-type compounds by their molecular structure is considered. Based on the synthesized compounds, a neuro-fuzzy model "structure - activity" was created. To predict the properties of these compounds, a numerical research method such as fuzzy logic was used and the possibility of using an expert model based on the Sugeno algorithm to determine the acute toxicity of 4-azatricyclo [4.2.1.03,7] nonanes derivatives by the structural formula was shown. The proposed methodology for creating an expert model can be applied to any classes of compounds, provided that appropriate descriptors are selected that characterize the molecular structure of the selected class of substances and that there are enough compounds to train a neuro-fuzzy hybrid network.
Key words
neuro fuzzy modeling, prediction, training sample, descriptor, acute toxicity.
7. Kryglov V.V. Nechetkaya logika i iskysstvennie neironnie seti / M.: Izdatelstvo Fiziko-matematicheskoi literatyri, 2001. 224 s.
8. Baskin I.I., Skvortsova M.I., Palyulin V.A., Zefirov N.S. Quantitative chemical structure - property/activity relationship studies using artificial neural networks. / Foundations of Computing and Decision Sciences, 1997. 22. 2. P. 107-116.
9. Cramer R.D., Patterson D.E., Bunce J.D. Comparative molecular field analysis (CoMFA). Effect of shape on binding of steroids to carrier proteins / Hi. Am.Chem. Soc, 1988. 110. P. 5959-5967.
References
1. Derbisher E.V., Vedenina N.V., Aleksandrina A.Yu., Derbisher V.E. Methodology for predicting the hazard class of little-known organic compounds / Modern high technology. 2007. 8. R. 60-62. [In Russian].
2. CambridgeSoft ChemBio Office Ultra 13.0 [Electronic resources]. http:// www.cambridgeSoft-.com / Date of acces: 21.08.2016.
3. Dragon 7 [Electronic resources]. http://www.talete.mi.it / Date of acces: 12.02.2020.
4. Todeschini R., Consonni V. Handbook of Molecular Descriptors / Weinheim: WILEY-WCH Publishers, 2000. 688 p.
5. Balaban A.T. From Chemical Topology to 3D Geometry / J.Chem.lnf.ComputSci, 1997. 37. P. 645-650.
6. Baskin I.I., Gal'bershtam N.M., Palyulin V.A. Computer implementation of artificial neural networks to solve the problems of revealing the connection «structure-property» / Information technology, 1997. 9. R. 27-30. [In Russian].
7. Kruglov V.V. Fuzzy logic and artificial neural networks / M.: Publishing house of physical-mathematical literature, 2001. 224 r. [In Russian].
8. Baskin I.I., Skvortsova M.I., Palyulin V.A., Zefirov N.S. Quantitative chemical structure - property/activity relationship studies using artificial neural networks. / Foundations of Computing and Decision Sciences, 1997. 22. 2. P. 107-116.
9. Cramer R.D., Patterson D.E., Bunce J.D. Comparative molecular field analysis (CoMFA). Effect of shape on binding of steroids to carrier proteins / Hi. Am.Chem. Soc, 1988. 110. P. 5959-5967.