The era of AI in drug discovery and development is just beginning, and industry insiders foresee a massive shift not only in speed and scale but also in what it takes to become a successful drugmaker. With advances in machine learning overlapping with new biological advances, the early work of AI integration will likely yield concrete benefits for patients within the next five years, according to Dr. Mikael Dolsten, former chief scientific officer at Pfizer.
Dolsten, who is now an adviser to the AI-based drug discovery and development company Immunai, envisions a future lab where science is integrated with medical practice almost in real-time. Industry successes like Insilico Medicine’s AI-designed pulmonary fibrosis candidate rentosertib and Google DeepMind’s AlphaFold platform for three-dimensional protein mapping are just the beginning. The industry is optimistic about solving some of the toughest challenges in drug discovery with AI.
One key aspect of AI in drug discovery is the abundance of data. AI thrives on scientific and clinical data, and companies like AbbVie have vast databases like their R&D search engine “ARCH,” which holds around 450 terabytes of data. Moving forward, researchers aim to mitigate errors and hallucinations that could potentially mislead development programs.
Stef van Grieken, CEO of the AI drug discovery software company Cradle, emphasizes the importance of AI platforms being introspective. The next phase of AI will require platforms to discern which outputs make sense and which do not. By improving the ability of AI models to know when they do not know something, researchers can avoid going down the wrong path.
Novo Nordisk has collaborated with Cradle to search for new cardiometabolic drug candidates that could improve upon existing blockbusters. Mishal Patel, senior vice president of AI and digital innovation at Novo, highlights the speed and cost-effectiveness of using AI libraries for drug discovery. The ability to sift through vast amounts of data quickly gives companies like Novo Nordisk a competitive advantage in the industry.
As AI continues to evolve in the field of drug discovery and development, understanding its limitations and leveraging its strengths will be crucial for success. The future holds much promise for AI in revolutionizing the way drugs are discovered, developed, and brought to market, ultimately benefiting patients worldwide. Navigating through the complexities of drug discovery and development can be likened to weaving your way through the woods on a bicycle – agile and fast decision-making is key. Just as Patel compared the search for novel targets to finding new stars in the galaxy, the use of AI in this field shines a powerful spotlight on areas of biology that may be poorly understood.
CEO of Immunai, Noam Solomon, emphasizes the importance of failure in machine learning, highlighting that success stories are only part of the process. Solomon’s company is mapping the immune system using AI platform Amica, collaborating with pharma partners like Pfizer to make critical decisions, even if it means recommending the termination of a clinical program.
Amica’s single-cell multiomic sequencing of human immune systems is a groundbreaking approach, with a database expected to reach 10 billion cells in five years. Solomon is also leading an academic collaboration effort to provide immunology researchers with free access to sequencing technology, aiming to fill gaps left by cuts to the NIH and enhance Amica’s datasets.
Former CSO of Pfizer, Dolsten, recognizes the value of public-private partnerships in driving innovation, especially in the context of technology, AI, and clinical science. The goal is to make significant advancements in knowledge by leveraging these collaborations.
The impact of AI on the pharmaceutical industry is still unfolding, with leaders like Novo’s Patel noting the shrinking gap between computational and traditional roles in research. AI developers are working closely with scientists to ensure a seamless convergence of expertise, leveraging tacit knowledge and providing value at scale.
Dolsten stresses the importance of balancing dry lab and wet lab resources, highlighting the shift towards modeling outcomes and reducing the reliance on traditional methods. Human intelligence remains crucial in the final product, underscoring the need for a harmonious blend of AI technology and scientific expertise.
As the industry continues to evolve, embracing AI as a powerful tool in drug discovery and development will be essential for driving innovation and advancing healthcare. With a collaborative approach and a focus on leveraging the best of both worlds, the lab of the future holds immense potential for groundbreaking discoveries and improved patient outcomes.
