Mapping the Landscape of AI and Machine Learning in Drug Formulation Design: A Bibliometric Analysis
Description
This bibliometric analysis explores AI and Machine Learning (ML) in drug formulation, particularly oral drug delivery. AI/ML addresses significant challenges in oral drug development, like solubility and bioavailability, enhancing efficiency in drug discovery. The study aimed to comprehensively overview the field by analyzing publication trends, identifying leading countries, institutions, and authors, and exploring co-citation networks to guide future research. Data was systematically collected from the Scopus database, focusing on English-language journal articles from 2023 to July 2025 concerning AI/ML in drug formulation. VOSviewer and other tools were used for data refinement and analysis, including publication trends, geographic contributions, and keyword co-occurrence. Results show robust publication growth, especially in 2024, indicating accelerating interest. The United States leads in research output, followed by China, India, and the UK, highlighting global collaboration. Influential publications underscore AI's practical applications in optimizing drug formulations. In conclusion, AI and ML are significantly advancing drug formulation, marked by increasing global engagement. Future research will likely focus on refining predictive models, optimizing manufacturing, and developing innovative AI-driven strategies for improved drug characteristics. The field continues to evolve through the synergy of technology and pharmaceutical sciences.
Files
Steps to reproduce
2.1 Data Source and Search Strategy This bibliometric analysis primarily utilized Scopus as the database for data retrieval, chosen for its comprehensive coverage of scientific literature in relevant fields like pharmaceutical sciences, chemistry, and engineering. The objective was to identify publications focusing on the application of artificial intelligence (AI) and machine learning (ML) in drug formulation design. The specific search string employed was: ( TITLE-ABS-KEY ( artificial intelligence ) AND TITLE-ABS-KEY ( drug formulation ) ) AND PUBYEAR > 2022 AND PUBYEAR < 2026 AND ( LIMIT-TO ( DOCTYPE , "ar" ) ) AND ( LIMIT-TO ( LANGUAGE , "English" ) ) AND ( LIMIT-TO ( SRCTYPE , "j" ) ) AND ( LIMIT-TO ( SUBJAREA , "CHEM" ) OR LIMIT-TO ( SUBJAREA , "MATE" ) OR LIMIT-TO ( SUBJAREA , "MULT" ) OR LIMIT-TO ( SUBJAREA , "ENGI" ) OR LIMIT-TO ( SUBJAREA , "CENG" ) OR LIMIT-TO ( SUBJAREA , "IMMU" ) OR LIMIT-TO ( SUBJAREA , "COMP" ) OR LIMIT-TO ( SUBJAREA , "DENT" ) OR LIMIT-TO ( SUBJAREA , "HEAL" ) OR LIMIT-TO ( SUBJAREA , "BIOC" ) OR LIMIT-TO ( SUBJAREA , "PHAR" ) OR LIMIT-TO ( SUBJAREA , "MEDI" ) ). This string was designed to capture articles with the specified keywords in their title, abstract, or keywords, published between 2023 and 2025 (specifically up to the end of July 2025). The search was limited to English-language journal articles within several subject areas, including Chemistry, Materials Science, Engineering, Computer Science, and Pharmacology, Toxicology and Pharmaceutics, among others. The initial search yielded 136 documents. 2.2 Data Collection and Refinement Following the initial data retrieval, a rigorous process of data collection and refinement was undertaken. Duplicate records were systematically removed to ensure the uniqueness of each entry, guided by principles like the PRISMA framework. Strict inclusion and exclusion criteria were applied based on the predefined search strategy, ensuring that only English-language journal articles within the specified subject areas were included in the final dataset. 2.3 Bibliometric Software and Tools For the comprehensive analysis of the collected data, several specialized bibliometric software and tools were employed. VOSviewer was utilized for network visualization, enabling the graphical representation of co-citation and co-occurrence networks. Harzing's Publish or Perish was used for extracting and analyzing citation metrics, while Microsoft Excel served for preliminary data cleaning and organization prior to more in-depth analyses. 2.4 Data Analysis Techniques The refined dataset was subjected to various bibliometric analysis techniques to address the study's objectives. Publication trends were analyzed by examining annual publications and cumulative growth to identify periods of rapid increase in research output. Geographic and institutional contributions were assessed by mapping the affiliations of authors to pinpoint leading countries and institutions.
Institutions
- Cyberjaya University College of Medical Sciences