ABSTRACT
Using the computer system SARD-21 (Structure Activity Relationship & Design) structural features of high- and low-effective anti-arrhythmic agents have been recognized and the influence of these features on the anti-arrhythmic properties has been evaluated. This information has been used for generation of the model for prediction of anti-arrhythmic effectiveness of pharmaceutical preparations with the 82%-level of recognition by two methods. The recognized structural parameters may be successfully used for design of new highly effective anti-arrhythmic drugs, and also for modification of structures of known anti-arrhythmic drugs for the increase of effectiveness of their anti-arrhythmic action.
Subject(s)
Amides/chemistry , Anti-Arrhythmia Agents/chemistry , Models, Theoretical , Quantitative Structure-Activity RelationshipABSTRACT
A relationship was studied between the structure, anthelmintic and herbicidal activities and toxicity of the derivatives of benzimidazole, sulfonyl urea, and hydrocarboxylic acids. Reliable mathematical recognition and prediction models were developed. Molecular design and prediction of potentially active structures with low toxicity and low-toxic compounds with a complex of properties (herbicidal, growth-regulating, and fungicidal activities) were accomplished.
Subject(s)
Drug Design , Herbicides/toxicity , Toxicology/methods , Toxicology/organization & administration , Animals , Guinea Pigs , Herbicides/chemistry , Mice , Mice, Inbred C57BL , Rabbits , RatsABSTRACT
Using the computer system SARD-21 (Structure Activity Relationship & Design) the structural features typical for high- and low- effective nonsteroid anti-inflammatory drugs (NSAIDs) were analyzed. This information has been used for the model for prediction of anti-inflammatory effectiveness of medicines with 76% and 81% level of recognition by two methods. New data can be used for creating new highly effective NSAIDs, and for increasing effectiveness of already known components.
Subject(s)
Anti-Inflammatory Agents, Non-Steroidal/chemistry , Computer Simulation , Cyclooxygenase Inhibitors/chemistry , Drug Design , Prostaglandin-Endoperoxide Synthases/chemistry , Molecular Structure , Software , Structure-Activity RelationshipABSTRACT
A method has been proposed to recognize and predict the range values of acute toxicity (LD50) and for this a system has been developed. The essence of the method is to gradually decrease the predictable toxicity ranges for the test substances in a stepwise manner while advancing within the formed predictive complexes. The transmission of information on the structure of compounds along this complex is determined by the preset logistic scheme along which narrower LD50 ranges are designated. The intervals are automatically set, by optimizing the recognition of the properties of compounds at the stage of formation of models (that recognize the set of signs). The predicting system includes two predicting complexes and contains a total of 30 mathematical models: for heterocyclic and carbocyclic compounds. Logistic schemes for passage of predictable compounds have been developed for each complex in accordance with the hierarchic organization of the elements that are constituents of the complex. The prediction significance is 70-90%.
