Development a novel robust method to enhance the solubility of Oxaprozin as nonsteroidal anti-inflammatory drug based on machine-learning

Abstract

Accurate specification of the drugs’ solubility is known as an important activity to appropriately manage the supercritical impregnation process. Over the last decades, the application of supercritical fluids (SCFs), mainly CO2, has found great interest as a promising solution to dominate the limitations of traditional methods including high toxicity, difficulty of control, high expense and low stability. Oxaprozin is an efficient off-patent nonsteroidal anti-inflammatory drug (NSAID), which is being extensively used for the pain management of patients suffering from chronic musculoskeletal disorders such as rheumatoid arthritis. In this paper, the prominent purpose of the authors is to predict and consequently optimize the solubility of Oxaprozin inside the CO2SCF. To do this, the authors employed two basic models and improved them with the Adaboost ensemble method. The base models include Gaussian process regression (GPR) and decision tree (DT). We optimized and evaluated the hyper-parameters of them using standard metrics. Boosted DT has an MAE error rate, an R2-score, and an MAPE of 6.806E-05, 0.980, and 4.511E-01, respectively. Also, boosted GPR has an R2-score of 0.998 and its MAPE error is 3.929E-02, and with MAE it has an error rate of 5.024E-06. So, boosted GPR was chosen as the best model, and the best values were: (T = 3.38E + 02, P = 4.0E + 02, Solubility = 0.001241).

Simulation and Optimization: A New Direction in Supercritical Technology Based Nanomedicine

“…Consequently, in numerous studies, researchers have resorted to utilizing classical ML algorithms to address tasks related to SCF-based pharmaceutical formulation preparation. Remarkably, these approaches have yielded commendable predictive results [65][66][67][68]. Conversely, employing models with a high number of parameters may lead to overfitting of the data, resulting in less-than-ideal predictive outcomes.…”

“…Each node in the tree represents a feature, each branch signifies a feature value, and the leaf nodes indicate the final prediction result. [21,65,68,69] Extra Trees In the context of decision tree-based ensemble algorithms, a random feature is chosen during node splitting. [63,66] Table 6.…”

Prediction of Drug-like Compounds Solubility in Supercritical Carbon Dioxide: A Comparative Study between Classical Density Functional Theory and Machine Learning Approaches

“…At present, one of the most promising approaches for such a task involves the use of machine learning (ML) methods, as evidenced by the recent growth in the literature dedicated to the topic. − However, it is important to note that a majority of the constructed QSPR models are mostly used for the inter- or extrapolation of data, based on a rather small initial set. Notably, there are two noteworthy works , which utilize a larger data set of over a hundred drugs for model training, claiming that their models can successfully predict the properties of new compounds, expanding beyond the original training set.…”

Computational intelligence modeling of hyoscine drug solubility and solvent density in supercritical processing: gradient boosting, extra trees, and random forest models

“…Either physical or chemical methods can be used for increasing the solubility of drugs in aqueous media, however the method of nanonization based on physical methods has attracted much attention recently for preparation of drug nanoparticles. One of the physical methods for drug nanonization is supercritical processing which can be used to prepare drug particles at nano size for enhanced aqueous solubility 2 . For developing this new technique, drug solubility in the supercritical solvent must be known prior to process design and development.…”

“…Estimating pharmaceutical solubility in supercritical solvents such as CO 2 has been reported by different methods such as thermodynamics and data-driven models 3 . The main inputs for the modeling have been considered to be pressure and temperature as these factors showed the most important effects on the drug solubility change 2 , 4 – 7 . It is a crucial step to measure and correlate drug solubility to prepare drugs with nanosized and better bioavailability.…”