Clarkston Consulting https://www.facebook.com/ClarkstonConsulting https://twitter.com/Clarkston_Inc https://www.linkedin.com/company/clarkston-consulting http://plus.google.com/112636148091952451172 https://www.youtube.com/user/ClarkstonInc
Skip to content

Applications of AI in Drug Development

Over the last two years, discussions about the applications of AI have broken through the mainstream and reached a level of hype in business circles that only tectonic shifts in business can attain. This uproar has left many organizations scrambling, from the board down, to determine specific ways to leverage AI to address challenges, improve business outcomes, and drive growth. This is especially true in the life sciences industry, where there’s no shortage of challenges facing the industry. Pricing pressures and the push to meet unmet patient needs have made efficient drug development more critical than ever before. Many biopharmaceutical companies are scrambling to utilize applications of AI to help overcome these challenges. Astellas, Takeda, and Pfizer are just a few examples of companies that are investing in the capabilities that AI can deliver in the drug development process.

With so many opportunities for improvement and so many challenges to tackle, companies have found it difficult to prioritize and agree on areas for investment. Yet a critical business process has risen to the top with significant agreement across the industry – drug development.

There are myriad ways forward-looking companies are innovating and leveraging AI-powered solutions into the drug development life cycle.

Using AI to Find Novel Applications for Established Medicines

Established and approved medicines in market already have a mountain of data about them available. As a result, finding new applications may represent a shortcut to creating new value. Astellas, for example, has partnered with Biovista, NuMedii and Excelra to use artificial intelligence to speed up this process. All three partners bring unique capabilities to this effort from an AI and big data perspective, including proprietary algorithms and annotated data feeding analysis of millions of clinical data points. These solutions have the capability to quickly analyze a tremendous amount of data to draw connections between existing compounds and disease targets.

Teva Pharmaceuticals is another example of a company looking to capitalize on the opportunity to leverage AI to repurpose drugs by partnering with IBM for use of their Watson platform. Watson has the capability to leverage a tremendous amount of medical journal articles and data to try to predict relationships between drug molecules and diseases. This technology allows scientists to more efficiently test hypothesis and consume the continuous stream of medical findings that could lead to the next big breakthrough for drug repurposing.

Using AI to Assist in the Identification of Net-New Compounds

Life sciences organizations already spend billions of dollars annually searching for the next compound that could ultimately turn into a life-saving drug. In an attempt to expedite this process and save on costs, companies are training AI to search massive data sets. Takeda, for example, is partnering with Numerate to identify clinical candidates for oncology, gastroenterology, and central nervous system disorders. Numerate’s artificial intelligence platform aims to dramatically improve efficiency in the small molecule drug discovery process. This partnership is beneficial to both parties as Numerate continues to validate and refine its AI solution, and Takeda is able to explore a greater number of new clinical candidates.

Technology companies are also making attempts to become a strategic partner for life sciences organizations in this space. Atomwise is an organization that has developed a proprietary AI platform, called AtomNet, to help scientists more efficiently identify drug candidates. Specifically, the technology focuses on the statistical likeliness that a small molecule will bind to proteins. Chemists typically have to test thousands of compounds to achieve the optimal combination of properties such as potency and limiting side effects. Using a computer to help with this optimization significantly speeds up the process and can lead to quicker development of critical drugs that patients need today.

Applications of AI in Trial Recruitment

The clinical trial recruitment process is another significant roadblock to getting a new drug approved. Pfizer, for example, worked with Alation Inc., Dataiku, and Tableau to develop an in-house analytics platform to help identify patients with rare diseases that may have gone undiagnosed. The tool, which took 2 years to build, is being used by Pfizer employees today to gather new insights that previously wouldn’t have been achievable. This technology has the potential to educate physicians about how to identify these rare diseases that can have symptoms that could easily be mistaken for more common diseases. It also provides Pfizer the opportunity to identify good patient candidates for clinical trials. This type of capability stands to increase in value as pharma and biotech continues exploring rare diseases, more complex generics, and new treatment methods.

Deep6 is another example of a company that has created an AI platform that uses natural language processing to mine through a tremendous amount of clinical trial data to find patients for clinical trials. Given the amount of unstructured data that currently exists for clinical trials, this solution meets a critical need for the industry to be able to convert patient data into a usable format. Overcoming the significant roadblock of recruiting patients is more important than ever before, as the pressure to deliver on unmet patient needs continues to mount.

Putting It All Together

As AI-enabled technology continues to improve, it will become more critical for organizations to pick the right areas to invest in the potential applications of AI for their business. Drug development is certainly a key area for opportunity – but that still leaves many questions unanswered.

Not only must the modern life sciences organization agree on where to apply their AI efforts, but they must also carefully determine how to ask the right questions of their data and pick the right set of strategic partners. Leaders need to build cross-functional teams internally to determine best use-cases and identify quick wins. Leaders must also be thinking about and planning for the future capabilities that their people will need to have in order to best leverage these technology tools to their full potential.

Coauthor and contributions by Raj Patel

Tags: Life Sciences Trends
RELATED INSIGHTS