Artificial intelligence (AI) has been evaluated as a tool to support various stages of drug development, from target discovery to adaptive clinical trial design. Now, this technology is offering tangible benefits for chemists involved in designing novel compounds or identifying new drug candidates. Progress in AI offers the exciting possibility of pairing it with cutting-edge lab automation, essentially automating the entire R&D process from molecular design to synthesis and testing – greatly expediting the drug development process.
It’s no surprise that scientists in pharma and biotech organizations are considering ways to increase efficiency. Getting a single drug to market takes an arduous 10 to 12 years, with an estimated price tag of nearly $2.9 billion.1 Last year, consulting firm Deloitte calculated that the return on pharma’s R&D investment had decreased to 1.8%, the lowest since the firm began evaluating it in 2010.2 These numbers have put tremendous pressure on stakeholders involved in drug discovery to operate differently, finding opportunities to break the trends of rising costs and longer development times.
Used together, AI and automation could be part of the solution.
In the drug discovery process, AI greatly increases the intellectual computing power of medicinal chemists. The primary advantage it offers is the ability to evaluate far more design parameters in parallel than a typical human brain can handle. When done well, the ultimate effect is to reduce the number of compounds that a scientist has to make and analyze in the lab to achieve the desired combination of physicochemical properties. In other words, AI enables drug discovery teams to be far more focused and efficient.
For example, for a typical drug program, getting to a single lead candidate can take three to five years and may involve the synthesis and analysis of as many as 2,000 to 3,000 molecules. With AI, scientists have been able to home in on a lead candidate from just 400 compounds.3
AI performs at its best when there is plenty of information available, such as in the IT or big data space where millions to billions of data points are often on hand. In drug discovery, we are lucky if we have a few hundred data points to start with – and AI does not work as effectively with such sparse data sets. This is where automation comes in.
The automation kicker
The benefits of automation in drug discovery are well known: increased reliability, throughput, and reproducibility, plus minimized hands-on time for tedious tasks. After all, various forms of automation have been a growing part of chemical synthesis workflows for more than two decades now, originating from the early days of combinatorial chemistry and more recently with the development of bench-scale flow chemistry systems.
Automated synthesis has traditionally focused on one- or two-step processes to make libraries of compounds for target screening and structure activity relationship development of increasing sophistication. However, cutting-edge technology is now enabling the fully automated multistep synthesis of quite complex molecules at scales from nanograms to grams, and at unprecedented speeds.4 For example, recent advances in inkjet technology have enable the “printing” of multistep reactions at a throughput of a reaction per second.5
This is where automation steps up to fill the sparse data problem in AI-guided molecular discovery. With the ability to rapidly make and test large numbers of targeted molecules, we can quickly fill the data gaps in AI models to predict molecular structures with desired properties.
For optimal utility, scientists should think of the AI–automation pairing as an iterative cycle rather than a one-step process. The more information fed into the AI, the better the output will be. Everything gleaned about building molecules through the automated workflow can be recorded and used to train the AI for the next cycle of experiments. By fully integrating both components into the drug discovery process, we have the potential for exponential impact in routinely reducing timelines for finding early drug candidates from years to a matter of months.
Put simply, AI streamlines the number of molecules that have to be synthesized, and automation makes it faster to build and test them. What this combination cannot do is replace the skill and expertise of trained and experienced scientists. AI and automation are best deployed to augment drug discovery chemists, allowing them to evaluate more possibilities more efficiently than can be done through the current state of the art. This approach allows drug discovery operations to be more nimble and efficient – chemists can run more programs simultaneously and make better decisions about which targets to move forward, getting more targets into the pipeline without a proportional increase in human effort. It can also enable teams to be more responsive to emerging diseases; indeed, scientists are already using this method to develop drugs for patients with COVID-19.6
Beyond that, the AI–automation pairing also stands to benefit downstream components as well, including process optimization for industrial chemistry and transferring existing molecules to automated manufacturing programs. Because these efforts are also very expensive with long timelines, they are big opportunities for efforts to reduce the time and money it takes to get a new drug to market.
Early implementation of AI for drug discovery has typically placed it in the hands of computational chemistry groups, where scientists already have the technical skills needed to integrate this new tool into molecule discovery. It is intriguing to consider that the development of more user-friendly – perhaps AI-driven – interfaces could expand access of sophisticated AI tools to a larger community of scientists who do not have the computational background but do know the properties of the molecules they need. With AI and automation, those opportunities may be on the horizon.
1. Dimasi, J. A., Grabowski, H. G., & Hansen, R. W. (2016). Innovation in the pharmaceutical industry: New estimates of R&D costs. Journal of Health Economics, 47, 20–33. doi: 10.1016/j.jhealeco.2016.01.012
2. Lesser, N., Shah, S. Deloitte Services LP. (2020, January 24). Measuring the Return from Pharmaceutical Innovation 2019. Retrieved from https://www2.deloitte.com/us/en/pages/life-sciences-and-health-care/articles/measuring-return-from-pharmaceutical-innovation.html
3. Mullard, A. (2017). The drug-makers guide to the galaxy. Nature, 549(7673), 445–447. doi: 10.1038/549445a
4. Gesmundo, N. J., Sauvagnat, B., Curran, P. J., Richards, M. P., Andrews, C. L., Dandliker, P. J., & Cernak, T. (2018). Nanoscale synthesis and affinity ranking. Nature, 557(7704), 228–232. doi: 10.1038/s41586-018-0056-8
5. White, J.D. SynJet: A Novel Chemical Dispensing Platform for High-throughput Reaction Screening and Optimization. September 30, 2019, Print4Fab conference, San Francisco, CA. Publisher: Society for Imaging Science and Technology.
6. Scudellari, M. (2020, March 19). Five Companies Using AI to Fight Coronavirus. IEEE Spectrum. https://spectrum.ieee.org/the-human-os/artificial-intelligence/medical-ai/companies-ai-coronavirus
Nathan Collins, Ph.D., is Chief Strategy Officer of SRI Biosciences, where he oversees the translation of R&D programs into commercially available platforms. A chemist by training, he spent years in drug discovery and is now focused on improving the synthetic chemistry process.