Emerging Models of AI Speed Drug Discovery
New models of collaboration can and do drive AI innovation in beneficial directions.
Everyone is concerned with the harmful effects and short-term profit-taking associated with artificial intelligence (AI). However, AI has been used for decades and is being employed constructively to solve complex problems that benefit society. Government, industry, and academia have established collaborative partnerships with incentives that strengthen positive outcomes. Such emerging ecosystems have already occurred in the pharmaceutical industry to speed up vaccine development and drug research in a responsible fashion.
Traditionally, drug manufacturers take ten to twelve years at a cost of tens of billions of dollars to discover a new drug for a targeted disease. Once developed, most drugs fail during the clinical trial evaluation stage. Digital platforms combined with AI are reducing the time and cost of initial drug discovery and may soon be used to run computer simulated models of early phase clinical trials.
Working with government agencies under Operation Warp Speed, Moderna and Pfizer developed their mRNA vaccines in eleven months instead of decades, thereby creating vaccine history. AI allowed for the rapid isolation and digital distribution of the SARS-CoV-2 genome on a global basis, which was shared with over fifty nations within a few short weeks. Until now, we have not seen such widespread speed in collaboration between countries and scientists.
The protein spike on the SARS-CoV-2 virus has a unique configuration that allows it to easily enter the cell. This underlying structure caused the virus to spread quickly during the early pandemic. Using AI, scientists produced, in record time, a mRNA vaccine by reverse-coding the original base amino acids within a specific portion of the protein spike. For their early groundbreaking research in mRNA technology at the University of Pennsylvania, biochemist Katalin Karikó and immunologist Drew Weissman won the Nobel Prize in Physiology and Medicine. Their work laid the foundation for developing the BioNTech/Pfizer and Moderna COVID-19 vaccines.
AI was also used to track the epidemiological spread of COVID-19 and to monitor the distribution of vaccines. Drug companies established early manufacturing processes for mRNA during the research and clinical trial phases rather than later in the process. The speed and scale at which AI can be utilized to achieve these goals helped reduce the timeframe for vaccine development.
Artificial intelligence is also being directed toward new drug discovery. Biotechnology companies are establishing 3-D computer libraries of known drugs, chemicals, proteins, biochemical disease pathways, and genetic structures obtained through open-source databases and medical journals. Drug research companies are using these digital databases to identify drug molecules that can target disease processes within days or weeks (not years).
Within three days of receiving the genomic structure of the virus and using 3-D AI molecular analysis, one company, Benevolent AI (based in the UK), identified a drug treatment, Baricitinib. This drug, already approved for rheumatoid arthritis, could be authorized for severe cases of COVID-19. The drug reduces the cellular viral load and blocks the cytokine cascade pathway, which causes overstimulation of the immune system leading to severe respiratory distress and organ failure.
In early 2020, Baricitinib was evaluated in early clinical trials for patients in Italy with severe COVID morbidity. Eventually, the drug was approved for widespread use in the EU. In June, 2020, the Food and Drug Administration (FDA) issued Emergency Use Authorization for the drug for those with severe illness, in combination with Remdesivir, a broad-spectrum anti-viral.
In the future, the convergence of AI and quantum computing will further accelerate drug research and development. Companies such as IBM, Google, and IonQ, have developed quantum computers that function at speeds thousands of times faster than traditional computers. These organizations are working to scale up the use of quantum technology, such as in the pharmaceutical industry.
Using digital cloud platforms, quantum computers are now accessible to industry, including start-ups, to create quantum algorithms, which will speed up new drug identification and improve investigations of disease pathology. AI and quantum computers may also be used to develop simulations of early phase clinical trials, which cannot be accomplished on today’s computers.
Collaborative networks using AI and quantum computing can ensure that society responsibly benefits from shared resources and knowledge. One such network is in Israel. AION Labs is a consortium of four multinational drug companies (Pfizer, Merck, Teva, and AstraZeneca), the Israeli Innovation Authority, the Israel Biotech Fund, and Amazon Web Services (AWS). This alliance is the basis for a start-up biopharmaceutical ecosystem in Rehovot, Israel.
Israel ranks first in the world for per capita dollars spent in global research and development. At AION Labs, the larger drug companies support the collaborative growth of start-up organizations in drug research and development. These start-up companies focus on the science of drug and disease pathway research as well as related clinical trials.
The larger companies share resources with the start-ups, such as scientific expertise, production and distribution capabilities, AI and digital technology, and access to venture capital. The computer cloud service, AWS, provides the AI backbone for 3-D digital libraries of molecules and algorithm platforms to support open-source drug and disease discovery. AWS can also bring quantum computing capabilities to accelerate research and to design simulated early phase in situ clinical trials.
The start-up companies within AION Labs can further sustain growth by retaining patents and licenses for their discoveries, while partnering with the larger organizations for global manufacturing and distribution. This creates an economic engine for the growth of biopharmaceutical start-ups and job creation. Smaller companies can grow without being swallowed up by larger biotech companies or spending time finding buyers to manufacture a new drug. Using AI and shared resources, the cost of drug development is lowered.
The AION Labs consortium will focus on personalized medicine using mRNA vaccines to fight cancer. The immune system normally destroys cancer cells throughout our body. Certain cancer cells hide specific receptors from the body’s immune system, leading to uncontrolled cellular growth. Using AI to create mRNA vaccines could boost the immune system against these target receptors and fight cancers in combination with other drugs.
Moderna already provides such a vaccine against melanoma. According to Jeffrey Weber MD, PhD at Perlmutter Cancer Center at NYU and the American Society of Clinical Oncology, the mRNA vaccine reduces the risk of cancer relapse over time by 44% when administered with the immunotherapy drug Keytruda (Merck).
Amazon’s new Braket program brings four quantum computer providers together on the AWS cloud platform for the AION drug companies to access. This will allow the start-ups to accelerate mRNA cancer research, stimulate broader drug research, and develop in situ clinical trial simulations.
AION Labs is an example of the Israeli government partnering with industry, technology, and finance to innovate responsibly with AI and quantum technology. The Israeli government sets the overall direction for economic development and subsidizes new companies with eighty per cent of their capital needs. Paid back by AION Labs’ successful start-ups, the government’s investment is returned over time. Offset by public subsidies, venture capital firms within the BioTech Fund can invest a smaller amount of funds. If a start-up is successful, the venture capital firms receive higher returns with lower risk.
AI in the pharmaceutical industry can achieve benefits for society and financial success on a global basis. Emerging collaborative ecosystems between government, industry, and academia can build an entrepreneurial culture of small businesses through sustainable long-term growth. This culture provides both direction and constraints to advance AI and quantum computing in a responsible fashion.
Cultural ethicist Faye Lincoln is author of the book Values That Shape the World: Ancient Precepts, Modern Concepts (Dialog Press, 2021). She analyzes the values-based implications of US and Middle East policies based on history, religion, economics, and public policy.