Digital Twin for Drug Discovery and Development—The Virtual Liver 1 Introduction The process of developing a drug is extremely research-intensive spanning efforts in biology, chemistry and manufacturing, while being characterised by a low probability of success. It is estimated that 50,000 hits are tested to achieve a successful drug. The odds of a drug molecule eventually reaching patients are so low that only one in 12 drug molecules that are tested on humans in clinical trials make it successfully to the market. Toxicity and lack of efficacy account for greater than 60% of all drug failures1. The traditional approach to drug discovery is akin to a pipeline with targets, leads and molecules each needing to meet pre-set success criteria to progress towards clinical development and ultimately the market. In such a brute force approach, making the right

Abstract

Digital twins are defined as digital replicas of processes, systems
or devices developed to foster deeper understanding and prediction.
While the concept of digital twins has largely been applied in the
manufacturing industry, one could conceive of a digital twin that integrates
information from diverse scientific and clinical sources to represent
the complex and dynamic relationships within biological networks.
Such an integrative system would allow one to gain a deeper understanding
of the biology and be used as a predictive framework to design
better drugs. The liver is a key organ in the body that is implicated in
various diseases and injuries leading to drug failures and withdrawals.
The study describes the development of a digital twin of the liver by integrating the knowledge and understanding gained by studying various
liver functions, diseases and the effect of drugs, using a mathematical
framework based on ordinary differential equations. This twin has been
shown to be effective in reproducing the normal liver function, evolution
of disease and the impact of treatment. Finally, a system that couples the
twin with experimental measurements has been demonstrated to provide
insights into drug-induced liver injury. The approach described in this
paper is fairly general and can be applied to other organs and biological
systems to develop drugs more efficiently and safely.