In 2018, long before the artificial intelligence (AI) bandwagon started careening down the mainstream, DeepMind, a subsidiary of Google parent Alphabet, developed a program named AlphaFold to predict protein structures faster and more accurately than biologists. Predicting protein shapes is a key aspect of drug development, disease treatment, and treatment research, but it is a traditionally long and arduous process. Faster, more affordable medication may soon be within reach thanks to AI automation across the entire drug development pipeline.
However, even this promised panacea had its Achilles’ heel. MIT researchers found that AlphaFold is only concretely useful in one step of drug discovery: modelling the structure of the protein. The system cannot model how a drug physically interacts with the protein. AlphaFold may not be a catch-all in drug discovery, but it started a conversation that is now funnelling millions of dollars in investment into key markets across Asia and beyond. According to Deep Pharma Intelligence, investments in AI-backed drug discovery have tripled over the last four years, reaching a staggering US$24.6 billion in 2022.
The post-COVID-19 factor
During the pandemic, economies worldwide relied on AI-based medication discovery rather than traditional vaccine detection processes, which take years to create and are equally expensive, contributing to the market’s growth. For example, Pfizer collaborated with AI businesses to develop COVID therapies, which were approved in less than two years, compared to the typical 10-year process.
Related: AI-powered growth in healthcare: early trends and learnings
By using big scientific databases, reviewing drug candidates in silico, and expediting high-content screening tests with automated data processing, AI has proven its mettle in cutting drug discovery costs and delays. Now, the industry is making strategic decisions to bounce back post-COVID-19 through a big R&D push to bring advanced and accurate AI software to the market.
The current drug discovery process involves a time-consuming and costly trial-and-error approach. It costs approximately US$1.3 billion and 10 years to bring a new therapeutic drug to market, and this cost is expected to rise.
Clinical trials also have a notoriously high rate of failure — over 90 per cent by some estimates — which means eliminating trial and error might save businesses a lot of money while getting drugs from the lab to market more quickly.
The potential revenue is enormous; financial analysts at Jeffries estimate that Takeda’s move may generate up to US$3.7 billion in annual sales. Morgan Stanley estimates that the next 10 years may spawn up to 50 new AI-driven medicines worth more than US$50 billion in sales.
AI is a long-term strategy
Huawei Cloud is adopting a long-term investment strategy in AI-assisted drug design. The Huawei Cloud Pangu Drug Molecule Model, developed with the Shanghai Institute of Materia Medica, helps pharmaceutical companies build small[1]molecule drugs. The model uses data from over 1.7 billion compounds to streamline the process for researchers to then run targeted experiments to verify efficacy.
“AI could effectively function as a virtual chemist, helping researchers design and identify novel molecules that are likely to interact with drug targets,” says Dr. Qiao Nan, Head of Huawei Cloud EI Health.
According to Qiao, AI could shrink R&D costs by up to 70 per cent while helping scientists discover novel lead compounds in months rather than years.
Related: The Future of Medicine: AI's Journey in Healthcare
“This would make more potential drug candidates available for clinical trials, lifting the overall success rate in what traditionally has been hit-or-miss process and increasing the odds that a new chemical compound will eventually become an effective, life-saving drug.”
On the back of this success, Huawei launched a unique AI-assisted commercial pharmaceutical SaaS platform in China to help companies reduce the costs of trial and error, while accelerating the discovery of lead compounds from several years to just one month. The SaaS platform is slated to expand internationally, starting with APAC, the Middle East, and further afield.
Building a strong AI talent pipeline
AI applications lower the R&D gap in the drug manufacturing process and aid in targeted medication manufacturing. As a result, biopharmaceutical companies are turning to AI to increase their market share. However, AI for drug discovery requires machines to replicate human intellect to address complex drug development difficulties — a tall order for any platform that is meant to be a tool. Its success depends on how and by whom it is used.
Currently, China is leading the global AI industry, housing over 60 per cent of big data experts across sectors. As more industry segments begin to rely on AI, players in healthcare will need to start preparing for a gap between supply and demand for talent. The alternative, as seen by incumbents including Sanofi, Merck, and GSK, is growth through collaboration or acquisition.
As the demand for messenger RNA (mRNA) vaccines in Southeast Asia increased during the pandemic, Singapore quickly established itself as a hub for AI, robotics, and manufacturing for leading pharmaceutical companies to open their regional headquarters. The small city-state is home to 30 contract manufacturing facilities, most of which are foreign-owned, if not partially backed.
According to GlobalData’s Contract Service Provider database, Merck currently owns three facilities in Singapore, and Novartis and GSK each own two facilities. “New opportunities will emerge as the biomanufacturing industry undergoes major changes brought about by the rapid pace of digitalisation, Industry 4.0, and the need for greater sustainability,” says Lim Keng Hui, Assistant Chief Executive of Singapore’s Science and Engineering Research Council, A*STAR.
Recently, Japan’s Takeda Pharmaceutical has partnered with AI tech startups and hired additional data scientists to address this. AI is part of Takeda’s long-term strategy to save money and time by speeding up the medication development process.
In May, the global pharma giant acquired US-based AI startup Nimbus Therapeutics for US$4 billion. The start-up used AI and machine learning algorithms to pick a compound to treat psoriasis out of thousands of other molecules. The experimental drug has already passed the first two phases of human trials, meaning it could be one of the first therapies discovered with AI if it passes the final trials this year. AI is shifting the drug discovery paradigm by extracting hidden patterns and evidence from vast biomedical data, dramatically improving the clinical trial process while mining old drugs for new applications. The result is expected to bring treatments to patients faster.
This article appears in Omnia Health magazine. Read the full issue online today.
Access the must-read eBook on AI here to discover perspectives on its transformative role in the healthcare industry.
