Synergies between data science methods and innovative drug delivery technologies

Most areas of science and technology and beyond are undergoing an almost unprecedented rate of change, driven largely by the rapid growth in automation and robotics, computational power, and AI and machine learning algorithms. Many areas of science and medicine have adopted these technologies or are on a steep learning curve to do so in the short to medium term. Drug delivery systems that are very important for optimising therapeutic efficacy, patient compliance, and amelioration of side-effects are similarly undergoing a quiet revolution in modalities. However, drug delivery systems are arguably lagging many other scientific and biomedical fields in applying informatics, physics-based computational design and simulation approaches, and AI and machine learning to design, optimisation, and simulation of drug delivery systems. Here I review studies in which selected computational methods have been employed for these purposes, aiming to highlight their potential to accelerate the provision of more effective drug delivery systems and to identify modalities in which the benefits of these computational methods have not been achieved at all, or at least sub-optimally. The aim is to focus on less well-addressed existing and emerging drug delivery systems and to provide a perspective on what needs to be done, what could be done better, and where the synergistic partnership between computational/AI methods and contemporary drug delivery system may lead in the future.