Explore the vast range of titles available.

Solubility determination of poorly water-soluble drugs is pivotal for formulation scientists when they want to develop a liquid formulation. Performing such a test with different ratios of cosolvents with water is time-consuming and costly. The scarcity of solubility data for poorly water-soluble drugs increases the importance of developing correlation and prediction equations for these mixtures. Therefore, the aim of the current research is to determine the solubility of acetylsalicylic acid in binary mixtures of ethanol+water at 25 and 37°C. Acetylsalicylic acid is non-stable in aqueous solutions and readily hydrolyze to salicylic acid. So, the solubility of acetylsalicylic acid is measured in ethanolic mixtures by HPLC to follow the concentration of produced salicylic acid as well. Moreover, the solubility of acetylsalicylic acid is modeled using different cosolvency equations. The measured solubility data were also predicted using PC-SAFT EOS model. DSC results ruled out any changes in the polymorphic form of acetylsalicylic acid after the solubility test, whereas XRPD results showed some changes in crystallinity of the precipitated acetylsalicylic acid after the solubility test. Fitting the solubility data to the different cosolvency models showed that the mean relative deviation percentage for the Jouyban-Acree model was less than 10.0% showing that this equation is able to obtain accurate solubility data for acetylsalicylic acid in mixtures of ethanol and water. Also, the predicted data with an average mean relative deviation percentage (MRD%) of less than 29.65% show the capability of the PC-SAFT model for predicting solubility data. A brief comparison of the solubilities of structurally related solutes to acetylsalicylic acid was also provided.

Protein solubility is significant in producing new soluble proteins that can reduce the cost of biocatalysts or therapeutic agents. Therefore, a computational model is highly desired to accurately predict protein solubility from the amino acid sequence. Many methods have been developed, but they are mostly based on the one-dimensional embedding of amino acids that is limited to catch spatially structural information. In this study, we have developed a new structure-aware method GraphSol to predict protein solubility by attentive graph convolutional network (GCN), where the protein topology attribute graph was constructed through predicted contact maps only from the sequence. GraphSol was shown to substantially outperform other sequence-based methods. The model was proven to be stable by consistent R 2 of 0.48 in both the cross-validation and independent test of the eSOL dataset. To our best knowledge, this is the first study to utilize the GCN for sequence-based protein solubility predictions. More importantly, this architecture could be easily extended to other protein prediction tasks requiring a raw protein sequence.

Rapid solvent selection is of great significance in chemistry. However, solubility prediction remains a crucial challenge. This study aimed to develop machine learning models that can accurately predict compound solubility in organic solvents. A dataset containing 5081 experimental temperature and solubility data of compounds in organic solvents was extracted and standardized. Molecular fingerprints were selected to characterize structural features. lightGBM was compared with deep learning and traditional machine learning (PLS, Ridge regression, kNN, DT, ET, RF, SVM) to develop models for predicting solubility in organic solvents at different temperatures. Compared to other models, lightGBM exhibited significantly better overall generalization (logS  ± 0.20). For unseen solutes, our model gave a prediction accuracy (logS  ± 0.59) close to the expected noise level of experimental solubility data. lightGBM revealed the physicochemical relationship between solubility and structural features. Our method enables rapid solvent screening in chemistry and may be applied to solubility prediction in other solvents.

The primary goal of this research is to apply bagging-based regression techniques to forecast the solubility of raloxifene and the density of carbon dioxide (CO₂). Bagging regression models were utilized, namely Bagging Bayesian Ridge Regression (BAG-BRR), Bagging Linear Regression (BAG-LR), and Bagging Polynomial Regression (BAG-PR). The hyperparameters of these models were tuned using the Tree-Based Parzen Estimators algorithm to achieve optimal performance. The results demonstrate the efficacy of the bagging regression models in predicting both the CO 2 density and the solubility of raloxifene. For the CO 2 density prediction, BAG-BRR achieved a coefficient of determination (CoD/R 2 ) of 0.83728, an RMSE of 6.0525E+01, and an AARD% of 1.16098E+01. BAG-LR attained a CoD of 0.85705, an RMSE of 5.8358E+01, and an AARD% of 1.11066E+01. BAG-PR exhibited superior performance with a CoD of 0.98559, an RMSE of 2.5934E+01, and an AARD% of 4.68598E+00. Similarly, for the solubility of raloxifene prediction, BAG-BRR achieved a CoD of 0.90615, an RMSE of 6.5797E−01, and an AARD% of 1.36868E+01. BAG-LR attained a CoD of 0.90002, an RMSE of 6.8669E−01, and an AARD% of 1.54778E+01. BAG-PR demonstrated outstanding performance with a CoD of 0.98565, an RMSE of 2.8158E−01, and an AARD% of 6.28460E+00. The findings highlight the potential of bagging regression models, particularly BAG-PR, for reliable and accurate predictions of CO 2 density and the solubility of raloxifene.

In this work, we have examined the predictive capability of a machine leaning model based on the XGBoost framework as regards the solubility of active pharmaceutical ingredient-like molecules in organic solvents over a wide range of temperatures. A total of 30 binary mixtures has been investigated. The dataset was divided in two sets, with one set for training, testing and validation including solubility data for four solute compounds (butyl paraben, fenofibrate, risperidone, fenoxycarb) consisting of a total of 224 data points, and the second set used for prediction consisting of the solubility data for butamben, with 50 data points in total. The calculated root mean square errors (RMSLE) for the calculated solubility (train, test, validation) were 0.05, 0.09, 0.13 and 0.15, respectively, while the average RMSLE for the predicted solubility of butamben was 0.41. A total of 10 descriptors were considered in this work, comprising parameters for solute (heat of fusion, melting temperature, heat capacity and Hansen solubility parameter), two descriptors representing the solvent (dielectric constant and boiling temperature) as well as temperature, and drug and solvent names. The temperature-dependence of solubility has been captured accurately by setting a constraint on the XGBoost algorithm. A comparison between the performance of the machine learning model proposed and evaluated in this work, and the semi-predictive Flory-Huggins and the temperature-dependent NRTL-SAC models on the other hand, shows that for all the studied compounds, the machine learning model can deliver significantly improved capability to model as well as predict solubility.

We present a systematic machine learning study of the solubility of diverse pharmaceutical acids in deep eutectic solvents (DESs). Using an automated Dual-Objective Optimization with Iterative feature pruning (DOO-IT) framework, we analyze a solubility dataset compiled from the literature for ten pharmaceutically important carboxylic acids and augment it with new measurements for mefenamic and niflumic acids in choline chloride- and menthol-based DESs, yielding N = 1020 data points. The data-driven multi-criterion measure is applied for final model selection among all collected accurate and parsimonious models. This three-step procedure enables extensive exploration of the model’s hyperspace and effective selection of models fulfilling notable accuracy, simplicity, and also persistency of the descriptors selected during model development. The dual-solution landscape clarifies the trade-off between complexity and cost in QSPR for DES systems and shows that physically meaningful energetic descriptors can replace or enhance explicit COSMO-RS predictions depending on the application.

We present 1,2,3-triazolium- and imidazolium-based ionic liquids (ILs) with aromatic anions as a new class of cellulose solvents. The two anions in our study, benzoate and salicylate, possess a lower basicity when compared to acetate and therefore should lead to a lower amount of N-heterocyclic carbenes (NHCs) in the ILs. We characterize their physicochemical properties and find that all of them are liquids at room temperature. By applying force field molecular dynamics (MD) simulations, we investigate the structure and dynamics of the liquids and find strong and long-lived hydrogen bonds, as well as significant π–π stacking between the aromatic anion and cation. Our ILs dissolve up to 8.5 wt.-% cellulose. Via NMR spectroscopy of the solution, we rule out chain degradation or derivatization, even after several weeks at elevated temperature. Based on our MD simulations, we estimate the enthalpy of solvation and derive a simple model for semi-quantitative prediction of cellulose solubility in ILs. With the help of Sankey diagrams, we illustrate the hydrogen bond network topology of the solutions, which is characterized by competing hydrogen bond donors and acceptors. The hydrogen bonds between cellulose and the anions possess average lifetimes in the nanosecond range, which is longer than found in common pure ILs.

This study aimed to provide a rational experimental design to collect a minimum number of experimental data points for a drug dissolved in a given binary solvent mixture at various temperatures, and to describe a computational procedure to predict the solubility of the drugs in any solvent composition and temperature of interest. We gathered available solubility data sets from papers published from 2012 to 2016 (56 data sets, 3488 data points totally). The mean percentage deviations (MPD) used to check the accuracy of predictions was calculated by Eq. 10. Fifty-six datasets were analyzed using 8 training data points which the overall MPD was calculated to be 15.5% ± 15.1%, and for 52 datasets after excluding 5 outlier sets was 12.1% ± 8.9%. The paired t test was conducted to compare the MPD values obtained from the models trained by 7 and 8 training data points and the reduction in prediction overall MPD (from 17.7% to 15.5%) was statistically significant (p < 0.04). To further reduction in MPD values, the computations were also conducted using 9 training data points, which did not reveal any significant difference comparing to the predictions using 8 training data points (p > 0.88). This observation revealed that the model adequately trained using 8 data points and could be used as a practical strategy for predicting the solubility of drugs in binary solvent mixtures at various temperatures with acceptable prediction error and using minimum experimental efforts. These sorts of predictions are highly in demand in the pharmaceutical industry.

Efficiently identifying suitable solvents for active pharmaceutical ingredients (APIs) is critical in drug formulation, yet the vast number of possible solvent-solute combinations presents a significant experimental challenge. This study addresses this by developing a robust machine learning (ML) model for accurately predicting the solubility of three phenolic acids (syringic, p-coumaric, and caffeic) in various deep eutectic solvents (DESs), integrating both experimental and computational investigations. Measured solubility data showed that the choline chloride combined with triethylene glycol in a 1:2 molar ratio was the most efficient system for the dissolution of the studied APIs. Different ML models, utilizing nu-Support Vector Regression (nuSVR) as the core regressor and based on descriptor sets derived from COSMO-RS (Conductor-like Screening Model for Real Solvents) computations, were systematically evaluated. A novel methodology termed DOO-IT (Dual-Objective Optimization with ITerative feature pruning) was employed to address the common challenges of model development with limited, high-value datasets. The final optimal 10-descriptor nuSVR model, selected from an exhaustive, multi-run search, demonstrated outstanding predictive power, offering a highly reliable computational tool for guiding experimental screening, significantly accelerating the exploration of DES-based formulations. This research also provides a strong foundation for future machine learning-guided discovery of chemicals, offering an effective and transferable framework for developing QSPR models for various chemical systems.

Solubility has been widely regarded as a fundamental property of small molecule drugs and drug candidates, as it has a profound impact on the crystallization process. Solubility prediction, as an alternative to experiments which can reduce waste and improve crystallization process efficiency, has attracted increasing attention. However, there are still many urgent challenges thus far. Herein we used seven descriptors based on understanding dissolution behavior to establish two solubility prediction models by machine learning algorithms. The solubility data of 120 active pharmaceutical ingredients (APIs) in ethanol were considered in the prediction models, which were constructed by random decision forests and artificial neural network with optimized data structure and model accuracy. Furthermore, a comparison with traditional prediction methods including the modified solubility equation and the quantitative structure-property relationships model was carried out. The highest accuracy shown by the testing set proves that the ML models have the best solubility prediction ability. Multiple linear regression and stepwise regression were used to further investigate the critical factor in determining solubility value. The results revealed that the API properties and the solute-solvent interaction both provide a nonnegligible contribution to the solubility value.