In silico models and AI for drug discovery and development

Drug development and discovery is a lengthy and costly process involving the systematic search for novel compounds with specific biological activity. The whole process typically involves four steps (early drug discovery, a pre-clinical stage, clinical phases, and regulatory approval) and can take well beyond 10 years with average costs around US$2 billion [1]. Additionally, less than 15% of drugs reaching the pre-clinical stage are later approved by a regulatory agency, and these are only a very small fraction of all drugs that are initially investigated. This highlights the staggering costs of bringing a new drug to market and the high failure rate of the traditional path of drug discovery and development.

In silico drug development

Modern advances in computer power, algorithms and software development, as well as the availability of computing resources and their decreasing costs, have contributed to the development of in silico methods to support and aid drug discovery and development throughout the pipeline, with a particular impact on the early stages of drug discovery. In the last few years, the computer-aided drug discovery market has grown substantially reaching US$2.9 billion globally in 2021 and is projected to reach US$7.5 billion by 2030, according to a recent report [2].

A plethora of computational tools are routinely used to identify and select therapeutically relevant targets, study the molecular basis of ligand–protein interactions, structurally characterise binding sites, develop target-specific compound libraries, model target proteins, identify hits by ligand- and structure-based virtual screening, estimate binding free energy, and optimise lead compounds. All of these tools can be used to rationalise and increase the efficiency, speed, and the cost-effectiveness of the drug discovery process [3].

AI in computational drug discovery

Artificial intelligence (AI) generally refers to the theory and development of computer systems able to simulate intelligent behaviour. The field encompasses a broad range of sub-domains and corresponding techniques among which machine learning is a fundamental concept. In the search for faster and more cost-effective solutions to drug discovery and development, AI becomes an appealing ally to aid and support medicinal chemists and drug designers to make the process more productive. 

Drug design is in fact inherently a complex multi-objective optimization process where the number of potential options to explore goes well beyond what is systematically possible with traditional high-throughput screening techniques [4]. Contrary to human capabilities, a core strength of AI is its ability to make good use of increasingly large volumes of data for decision-making, highlighting patterns of potential interest and providing ways to prioritise tasks. Moreover, the latest developments in AI have the potential to significantly impact not only drug design but multiple sub-fields of the pharmaceutical industry, from drug discovery to pharmaceutical product management (including clinical trial design and monitoring, quality assessment and control, and pharmaceutical manufacturing) [5].  

Future perspectives 

While there are certainly challenges involved with computer-aided approaches (ranging from strategic and methodological tasks to human interaction-related hurdles) [6], in silico drug development and discovery is playing an ever increasing and critical role in the pharmaceutical sector and this is likely to increase in the near future. These computational methods are key for limiting the use of animal models in pharmacological research, for aiding the rational design of novel and safe drug candidates, and for repositioning marketed drugs, supporting medicinal chemists and pharmacologists during the drug discovery trajectory [7].

References

[1] S. Lim, The process and costs of drug development (2022). https://ftloscience.com/process-costs-drug-development/ accessed on Nov 3, 2022

[2] Computer-aided drug discovery market research, 2030. Allied market research https://www.alliedmarketresearch.com/computer-aided-drug-discovery-market-A16823 accessed on Nov 3, 2022

[3] In silico drug discovery and design. Theory, methods, challenges, and applications. (2016) Edited by C.N. Cavasotto, CRC Press. 

[4] P. Schneider et al. Rethinking drug design in the artificial intelligence era. Nature Reviews Drug Discovery (2020) 19, 353-364.

[5] P. Debleena et al. Artificial intelligence in drug discovery and development. Drug Discovery Today (2021) 26(1), 80-93.

[6] J.L. Medina-Franco. Grand challenges of computer-aided drug design: The road ahead. Frontiers in Drug Discovery (2021) 1:728551

[7] S. Brogi et al. Editorial: In silico methods for drug design and discovery. Frontiers in Chemistry (2020) 8:612.