Artificial Intelligence and Drug Discovery: The Companies Leading The Way
While the process of drug development can be defined in five steps, the time taken to cover these steps in the real world spans a decade, and sometimes more. The journey of new medicine from a laboratory to a pharmacy store is lengthy and complex, with costs running into billions of dollars. While pharmaceutical companies have achieved many breakthroughs, the returns on investment in R&D for the top global pharmaceutical companies continue to fall. This is why many pharmaceutical companies are exploring advanced technologies, such as artificial intelligence (AI) and machine learning, to reduce both time and costs involved in the process of drug discovery.
Here is an overview of the work and industry collaborations in 2019 till now.
Drug Discovery & Artificial Intelligence
The Pharmaceutical Research and Manufacturers of America (PhRMA) estimates “the average cost to research and develop each successful drug to be $2.6 billion. This number incorporates the cost of failures—of the thousands and sometimes millions of compounds that may be screened and assessed early in the R&D process, only a few of which will ultimately receive approval. The overall probability of clinical success (the likelihood that a drug entering clinical testing will eventually be approved) is estimated to be less than 12%.”
Going by the standard approach, the process of discovering a new drug requires data scientists to comb through humongous amounts of data to identify the right chemical compounds that could be tested for a targeted disease. Typically, cells representing a specific disease are exposed to a variety of compounds, and a microscopy snapshot is taken of each reaction that follows, running into millions.
Such a massive amount of data collected from each such experiment can become much more useful if meaningful information can be extracted out of it. This is possible using AI. While scientists have been using AI in recent years to sort through such data, it is now being applied more commonly and for more advanced studies and results.
A study by Janssen Research & Development (JNJ arm) concludes that the “AI method to be up to 250 times more efficient than the traditional method of drug discovery.” AI holds the potential to reduce timelines for drug discovery, increase accuracy of predictions on efficacy and safety as well as bring in better opportunity to diversify drug pipelines. A report by BCG estimates that the total spending on AI-related drug discovery and development tools in 2022 is expected to hit $1.3 billion.
Collaborations & Work So Far
“Insights gained through AI can potentially help researchers better understand new therapeutic targets and more rapidly identify viable drug candidates. Additionally, AI may reduce the time and costs associated with drug development and answer key translational questions,” according to Bristol-Myers Squibb.
In October 2019, Novartis (NVS) entered into a multi-year alliance with Microsoft (MSFT) to leverage data and AI to transform how medicines are discovered, developed, and commercialized. Additionally, Novartis is setting up an AI Innovation Lab with Microsoft as its strategic AI and data science partner.
Gilead Sciences (GILD) and insitro joined hands in April 2019 to discover and develop treatments for patients with nonalcoholic steatohepatitis (NASH). The insitro Human (ISH) platform applies machine learning, human genetics, and functional genomics to drive therapeutic discovery and development.
In January 2020, Bayer (BAYRY) and Schrödinger (SDGR) announced a five-year technology alliance to develop a comprehensive de novo design solution to accelerate the discovery of innovative, high-quality drugs. Schrödinger is using Google Cloud for its physics-based computational platform for drug discovery.
Exscientia, a UK-based AI-driven drug discovery company, has collaborations on various projects with some of the biggest companies. Bayer and Exscientia are leveraging the potential of AI in cardiovascular and oncology drug discovery. The collaboration focuses on early-stage research by using an AI-based algorithm to predict potential drug molecules. Celgene and Exscientia are working on accelerating the discovery of small molecule therapeutic drug candidates in oncology and autoimmunity. In January 2019, Exscientia entered a new collaboration with Roche to apply its AI drug discovery platform to design pre-clinical drug candidates in oncology.
The German pharma company Boehringer-Ingelheim (BI) is leveraging AI software platform Sherpa by startup Kairntech on tasks to better utilize the existing unstructured text-based information.
Pfizer (PFE) and Insilico are researching to explore new data and AI systems for potential therapeutic targets implicated in a variety of diseases. In September 2019, Insilico unveiled GENTRL, an AI system for drug discovery developed in collaboration with WuXi AppTec and Alán Aspuru-Guzik.
Bristol-Myers Squibb (BMY) and Concerto HealthAI entered into an alliance to broaden uses of real-world evidence and accelerate precision oncology innovations. In mid-2019, Eli Lilly (LLY) and Atomwise entered a multi-year agreement to apply Atomwise’s patented AI technology in support of Lilly’s preclinical drug discovery efforts.
Companies such as Amgen (AMGN), AstraZeneca (AZN), BASF (BASFY), Bayer, GlaxoSmithKline (GSK), LEO, Johnson and Johnson (JNJ), Eli Lilly, Merck (MRK), Novartis, Pfizer, Sunovion, Syngenta, and Wuxi are part of the Machine Learning for Pharmaceutical Discovery and Synthesis Consortium (MLDPS). MLDPS is a collaboration between the pharmaceutical and biotechnology companies and Massachusetts Institute of Technology to work towards the design of useful software for the automation of small molecule discovery and synthesis.
In January 2020, Sumitomo Dainippon Pharma and Exscientia announced that a new drug candidate created using AI for the treatment of obsessive-compulsive disorder has begun its critical trial.
The pharmaceutical world is increasingly engaging with technology companies to shorten the time to identify new drugs and repurpose current drugs. However, challenges remain, and much more needs to be done before AI can be truly adopted as a part of the drug discovery process. This process has begun, and holds promise.
Disclaimer: The author has no position in any stocks mentioned. Investors should consider the above information not as a de facto recommendation, but as an idea for further consideration. The report has been carefully prepared, and any exclusions or errors in reporting are unintentional.
The views and opinions expressed herein are the views and opinions of the author and do not necessarily reflect those of Nasdaq, Inc.