DE-INTERACT: A machine-learning-based predictive tool for the drug-excipient interaction study during product development-Validation through Paracetamol and Vanillin as a case study

Tailoring drug release in bilayer tablets through droplet deposition modeling and injection molding

This study explores the innovative production of personalized bilayer tablets, integrating two advanced manufacturing techniques: Droplet Deposition Modeling (DDM) and Injection Molding (IM). Unlike traditional methods limited to customizing dense bilayer medicines, our approach uses Additive Manufacturing (AM) to effectively adjust drug release profiles. Focusing on Caffeine and Paracetamol, we found successful processing for both DDM and IM using Caffeine formulation. The high viscosity of Paracetamol formulation posed challenges during DDM processing. Integrating Paracetamol formulation for the over-molding process proved effective, demonstrating IM's versatility in handling complex formulations. Varying infill percentages in DDM tablets led to distinct porosities affecting diverse drug release profiles in DDM-fabricated tablets. In contrast, tablets with high-density structures formed through the over-molding process displayed slower and more uniform release patterns. Combining DDM and IM techniques allows for overcoming the inherent limitations of each technique independently, enabling the production of bilayer tablets with customizable drug release profiles. The study's results offer promising insights into the future of personalized medicine, suggesting new pathways for the development of customized oral dosage forms.

Towards safer and efficient formulations: Machine learning approaches to predict drug-excipient compatibility

Predicting drug-excipient compatibility is a critical aspect of pharmaceutical formulation design. In this study, we introduced an innovative approach that leverages machine learning techniques to improve the accuracy of drug-excipient compatibility predictions. Mol2vec and 2D molecular descriptors combined with the stacking technique were used to improve the performance of the model. This approach achieved a significant advancement in the predictive capacity as demonstrated by the accuracy, precision, recall, AUC, and MCC of 0.98, 0.87, 0.88, 0.93 and 0.86, respectively. Using the DE-INTERACT model as the benchmark, our stacking model could remarkably detect drug-excipient incompatibility in 10/12 tested cases, while DE-INTERACT managed to recognize only 3 out of 12 incompatibility cases in the validation experiments. To ensure user accessibility, the trained model was deployed to a user-friendly web platform (URL: https://decompatibility.streamlit.app/). This interactive interface accommodated inputs through various types, including names, PubChem CID, or SMILES strings. It promptly generated compatibility predictions alongside corresponding probability scores. However, the continual refinement of model performance is crucial before applying this model in practice.