The use of artificial intelligence (AI) in drug development is revolutionizing the pharmaceutical industry and holding the promise of faster, more efficient, and cost-effective drug discovery. According to Niven R. Narain, Ph.D., President & CEO of BPGbio, the advancement of AI and computing is dynamically changing the landscape of drug development, ushering in a new era of innovation.
At the renowned J.P. Morgan Healthcare Conference in January 2024, one of the key topics of discussion was the application of AI to accelerate health innovation. The conference revealed a significant increase in capital invested in AI-driven pharma companies, reaching $59.3 billion as of December 2022—an impressive 27-fold rise since 2015—according to data from Deep Pharma Intelligence.
The application of innovative AI approaches is not limited to the biotech sector. Traditional big pharma companies are also actively involved in AI-driven drug discovery, either through partnering with small AI biotech companies or developing in-house AI units. For example, Pfizer collaborated with IBM's Watson to expedite drug discovery in immuno-oncology, and Sanofi engaged with Exscientia to leverage AI in identifying drug targets for metabolic diseases.
AI companies are revolutionizing the drug discovery process by leveraging existing data from previous experiments, robotics, and images instead of relying solely on traditional lab-based scientific research approaches. These AI-driven discoveries have garnered excitement within the industry and provided early evidence of the potential of AI in biopharmaceutical innovation.
However, while the way drugs are discovered has evolved with the use of technology and computing, the ultimate measure of success remains the same: successful clinical trials leading to FDA approvals. The industry is currently seeking tangible success metrics to prove the utility of AI in drug development. The rhetoric at the J.P. Morgan conference acknowledged the promise of AI but cautioned against unrealistic expectations.
The past 18 months have served as a cautionary tale on the limitations of AI capabilities in drug discovery, with some AI-designed drug candidates failing in clinical trials. Industry leaders now recognize that AI-based drug discovery without patient biology, such as samples and wet lab experiments, is risky and unlikely to be successful on its own.
As we enter 2024, there is an opportunity for companies that prioritize biology-first approaches to AI to make progress in clinical trials with drug candidates developed with the help of proprietary AI platforms. These companies can leverage AI algorithms to define target populations and ultimately shift the narrative from tech hype to real value for patients.
The winners in AI-powered drug development must acknowledge several truths about the potential of AI to improve drug discovery and development. Companies that effectively leverage AI are beginning to witness success in the clinic, while those who have not incorporated AI well are experiencing clinical failures. Savvy biopharma companies utilize AI to navigate vast amounts of patient data developed from real biological samples, rather than relying solely on public databases.
By leveraging real biology, AI platforms can efficiently identify and validate promising drug targets, discover methods for targeting complex biological pathways, and optimize later-stage clinical trials based on the characteristics of responders in earlier stage trials. This synergistic relationship between AI and biology has the potential to reshape the drug discovery landscape, combining data analysis with biological insights.
AI-driven drug discovery starts with the identification and validation of drug targets, similar to uncovering hidden treasures within a complex biological landscape. However, it doesn't end there. Precision in aligning biological profiles with clinically relevant patient data is crucial, much like decoding individual genetic patterns that shape a person's health journey. AI can play a significant role in refining this understanding.
Nevertheless, effective drug discovery still relies on a solid foundation of robust biological understanding. The integration of AI's computational power with biology acts as a catalyst, enabling the identification and validation of potential drug targets that are crucial for groundbreaking therapeutic advancements. It can also be used to better understand the characteristics of the individuals who will respond to the therapies being tested. Leveraging AI models in combination with real biological observations represents the next frontier in applying AI to drug development.
The landscape of AI in drug discovery encompasses various types, including machine learning, neural networks, and Bayesian AI, among others. According to Dr. Narain, a biology-first approach, utilizing Bayesian AI, offers hypothesis-free discovery and has the potential to redefine how drugs are conceptualized, discovered, and developed. Neural AI, on the other hand, helps decode the intricate relationships between genetic factors and common diseases, assisting in crucial decision-making within drug development pathways. Different AI modules should be utilized in different aspects of discovery to ensure effectiveness, as there is no one-size-fits-all approach.
The use of AI revolutionizes drug development by boosting efficiency, improving data analysis, and reshaping trial structures. This addresses the escalating costs and high failure rates that have plagued the industry. A biology-first AI approach enhances patient specificity, allows for quicker identification of viable candidates for clinical trials, and paves the way for faster success.
After years of both successes and setbacks, 2024 marks the beginning of a phase focused on tangible outcomes. AI-developed drugs and diagnostics are being validated through emerging clinical data, moving beyond the era of mere hype. The convergence of AI and biology has the potential to transform the pharmaceutical industry and bring forth groundbreaking advancements that benefit patients worldwide.
