Artificial Intelligence and its Growing Influence in the Pharmaceutical World
Artificial intelligence (AI) has made headway across almost every industry, automating complex tasks and increasing efficiency in the modern world. Likewise, it’s no surprise that AI has found a place within the pharmaceutical industry. In pharma, AI utilizes algorithms to mimic responsibilities conventionally performed by people that require human intelligence. This automation has led to breakthroughs in how we diagnose, treat, and prevent disease. This article will discuss how exactly the pharmaceutical industry uses AI to drive science and innovation within the drug development space.
AI in drug discovery
Drug development is a time-consuming and expensive process with a bleak success rate. In fact, a Massachusetts Institute of Technology (MIT) study found that only 13.8 percent of drugs in development get approved, which is greater than the 10.4 percent reported by Hay et al. and the 9.6 percent reported by Thomas et al. AI provides an opportunity for pharmaceutical companies to increase these odds by playing an important role in drug discovery1,2,3.
Traditional drug discovery methods are lengthy and costly, but AI can mitigate these issues in several ways. AI can help inform drug design by predicting the structure and shape of drug targets, drug-protein interactions, and drug activity. Additionally, AI can assist in screening drug candidates through prediction of a drug’s toxicity and activity profile. Understanding these factors can aid in more careful selection of validated molecules, resulting in a higher chance of clinical success4.
AI in manufacturing
Manufacturing is a large part of drug development. These processes have become increasingly complex and health agencies have become more rigid in their expectations of quality. AI can automate these processes, thereby improving workflow and ensuring consistency and quality of product 4.
AI in clinical trials
Clinical trials are one of the most expensive and time-consuming aspects of drug development, yet they are necessary to proving the safety and effectiveness of a drug candidate. Likewise, it can take upwards of 10 to 15 years to complete all three phases of clinical studies.
In terms of clinical trial design, AI is particularly useful in the realm of real-world data (RWD) and real-world evidence (RWE). These terms refer to the plethora of scientific, medical, and research data available in the real world. Oftentimes this data is derived from things such as electronic medical records. AI can help to pull, consolidate, and analyze this data to characterize disease and inform study design, efficacy thresholds, and safety parameters.
An example of how companies leverage AI in clinical studies involves the use of RWE in clinical trial design. Many investigational programs have begun constructing historical control arms using RWE. Doing such eliminates the need for a placebo arm in clinical trials, eliminating the need for patient enrollment in a control arm. This can help to save time, money, and resources for the pharmaceutical company, thereby increasing the efficiency in which drugs get approved.
Pharma can greatly benefit from implementing AI across the industry. Doing so will not only optimize drug development programs but also get drug to patients faster.
- Wong, C. H., Siah, K. W., & Lo, A. W. (2019). Estimation of clinical trial success rates and related parameters. Biostatistics, 20(2), 273–286. https://doi.org/10.1093/biostatistics/kxx069
- Hay, M., Thomas, D. W., Craighead, J. L., Economides, C., & Rosenthal, J. (2014). Clinical development success rates for investigational drugs. Nature Biotechnology, 32(1), 40–51. https://doi.org/10.1038/nbt.2786
- Thomas D. W, Burns J., Audette J., Carrol A., Dow-Hygelund C. and Hay M. (2016). Clinical Development Success Rates 2006–2015. San Diego: Biomedtracker.
- Paul, D., Sanap, G., Shenoy, S., Kalyane, D., Kalia, K., & Tekade, R. K. (2021). Artificial intelligence in drug discovery and development. Drug Discovery Today, 26(1), 80–93. https://doi.org/10.1016/j.drudis.2020.10.010