Machine learning based simulation of an anti-cancer drug (busulfan) solubility in supercritical carbon dioxide: ANFIS model and experimental validation

In this work, a novel machine learning method (MLM) was developed based on Neuro fuzzy system, i.e., ANFIS (Adaptive neuro fuzzy inference system). This neuro fuzzy algorithm was implemented for simulating the solubility of pharmaceuticals in supercritical carbon dioxide. The model of drug was busulfan which is an anti-cancer drug with large applications for patients. The drug efficacy can be improved by production at nanosize using supercritical technology. The processing parameters including pressure and temperature were evaluated in the process and were taken into account in the modeling of supercritical system. Also, the simulated output which is the only predicted variable was the solubility of drug in the solvent. Several collected data from literature were used to train the ANFIS structure, and the trained model was then used for testing step. Both training and testing steps in developing ANFIS model indicated great accuracy and very close to the experimental values. Indeed, 32 data points were used for the ANFIS modeling among which 12 were used in testing the ANFIS model. In developing ANFIS model, Trimf membership function was used for generation of FIS structure. The Fuzzy system was generated using the Grid Partitioning technique which indicated great accuracy in prediction of busulfan solubility data. The model developed in this work indicated that the ANFIS model can be used to accurately predict the solubility of drugs in supercritical solvents which can be consequently used for production of drugs with improved efficacy. (C) 2021 Elsevier B.V. All rights reserved.

Automated summary

In this study, a novel machine learning method based on ANFIS was developed for predicting drug solubility in supercritical solvents. The model showed high accuracy in both training and testing steps, indicating its potential for accurately predicting drug solubility and improving drug efficacy.