Predicting complexation performance between cyclodextrins and guest molecules by integrated machine learning and molecular modeling techniques

Abstract

Most pharmaceutical formulation developments are complex and ideal formulations are generally obtained after extensive experimentation. Machine learning is increasingly advancing many aspects in modern society and has achieved significant success in multiple subjects. Current research demonstrated that machine learning can be adopted to build up high-accurate predictive models in drugs/cyclodextrins (CDs) systems. Molecular descriptors of compounds and experimental conditions were employed as inputs, while complexation free energy as outputs. Results showed that the light gradient boosting machine provided significantly improved predictive performance over random forest and deep learning. The mean absolute error was 1.38 kJ/mol and squared correlation coefficient was 0.86. The evaluation of relative importance of molecular descriptors further demonstrated the key factors affecting molecular interactions in drugs/CD systems. In the specific ketoprofen–CD systems, machine learning model showed better predictive performance than molecular modeling calculation, while molecular simulation could provide structural, dynamic and energetic information. The integration of machine learning and molecular simulation could produce synergistic effect for interpreting and predicting pharmaceutical formulations. In conclusion, the developed predictive models were able to quickly and accurately predict the solubilizing capacity of CD systems. Current research has taken an important step toward the application of machine learning in pharmaceutical formulation design.

Identifying sarcopenia in advanced non‐small cell lung cancer patients using skeletal muscle CT radiomics and machine learning

“…[23][24][25][26][27] Nevertheless, the sensitivity to overfitting presents the toughest challenge to the LightGBM algorithm, particularly with regard to the small dataset. 28 Thus, it is necessary to carefully tune the parameters of the LightGBM model.…”

Computer-Aided Discovery of New Solubility-Enhancing Drug Delivery System

“…The curated dataset consisted of diverse API and non-API chemical structures in systems with 16 CDs of natural and semi-synthetic origin. Previously, to model guest-CD systems, an effort was made to curate the data originating from the distinct literature studies [ 35 , 36 ], including a large library of both guest molecules and CDs [ 17 ]. We have followed the best practices of QSAR modeling [ 31 ] to perform our study and assure of its reproducibility.…”

“…Their model employed data on consistently measured ΔG of formation for βCD systems, and the random forest method achieving the external validation R 2 = 0.66. Zhao et al [ 17 ] compared gradient boosting and deep neural network models for predicting ΔG of complex formation. Gradient boosting showed good predictive performance indicated by consistent in-sample and out-of-sample scores R 2 = 0.86, while DNN resulted in a model with R 2 = 0.76 and R 2 = 0.62 for in-sample and out-of-sample predictions, respectively.…”

Digital Pharmaceutical Sciences

“…Over 30 numerous descriptors related to the guest molecule, CD, and experimental conditions have been implemented in designing the machine learning models. LightGBM showed better prediction performance compared to the other models including RF and DL (33). Gao et al (2020) also implemented the lightGBM method for prediction of complexation performance of 341 drugs/phospholipid complex formulations described by over 40 molecular descriptors related to the properties of the drugs, solvents, and experimental conditions.…”

“…Comparison of Different Machine Learning Methods Commonly Used in Pharmaceutical Research*(12,17,(25)(26)(27)(28)(29)(30)(31)(32)(33) …”