Artificial neural network designed to identify NBOMe hallucinogens based on 3D- MoRSE descriptors and topological descriptors
Adelina Ion, S. Gosav, Mirela Praisler, Cătălin Negoiţă
International Conference on System Theory, Control and Computing October 1, 2019 DOI: 10.1109/icstcc.2019.8885908 (opens in new tab)
Study at a glance
AI-extracted from the abstract| Characteristics | Theoretical or computational study Peer reviewed |
|---|---|
| Sample size | 160 |
| Population | Illicit drugs including NBOMe hallucinogens, sympathomimetic amines, narcotics, and analgesics |
| Key points | An Artificial Neural Network using 3D-MoRSE and topological descriptors can classify NBOMe hallucinogens among a set of 160 illicit drugs. |
Abstract
Trafficking of synthetic hallucinogens remains one of the most serious threats to public health and safety, both globally and nationally. A novel group of toxic phenethylamine derivatives, referred to as NBOMe, recently gained prominence on the illicit market. 25I-NBOMe is a potent synthetic derivative of 2,5-dimethoxy-4-iodophenetylamine (2C-I), which is a classical hallucinogen with serotonergic action. In this paper, we are presenting an Artificial Neural Network (ANN) designed to identify NBOMe hallucinogens based on their 3D-MoRSE and topological descriptors. The database contains 160 illicit drugs such as NBOMe hallucinogens, sympathomimetic amines, narcotics and other potent analgesics, as well as their main precursors. The molecular structures of all these compounds have been first optimized and then the 3D-MoRSE descriptors and topological descriptors have been determined by using the Dragon 5.5 software. Several figures of merit have been calculated in order to evaluate the efficiency of the ANN system in identifying the class identity of an unknown sample. A comparative analysis of the results obtained with the 3D-MoRSE descriptors and with the topological descriptors is presented in detail.