Immune Network Technology on the Basis of Random Forest Algorithm for Computer-Aided Drug Design

摘要

The article is devoted to immune network technology of new drugs sulfonamides properties prediction based on chemical structural information processing using the descriptor approach. Nowadays, the establishment of effective methods of QSAR (Quantitative Structure Activity Relationships) to predict the properties of new chemical compounds and directional computational molecular design of drug compounds are the most important and urgent tasks of bioinformatics. The article proposes a technology based on immune network modeling to determine the substance of new drug compounds. There was presented the algorithm for the modeling of the dependences "structure-property" on the example of sulfonamides with different duration of action based on the biological approach of artificial immune systems. Sulfonamides have been classified according to prognostic groups with different durations of action. As a method of selection of informative descriptors there was used Random Forest algorithm. The simulation results are presented in the software WEKA (Waikato Environment for Knowledge Analysis) and R Studio.