Matej Mikulic
Research expert covering health, pharma & medtech
AI in the pharmaceutical industry – statistics & facts
In public and media, 2023 was undoubtedly a breakthrough year for artificial intelligence (AI). However, not many people would probably think of the biopharmaceutical industry in general and drug discovery specifically when asked about AI. This is even more interesting since most of the people had used certain biopharmaceutical products during the last three years in whose development AI already played a significant role: the COVID-19 vaccines, especially those by Moderna and Pfizer/Biontech.
Early R&D stages are in focus
AI’s potential for reshaping and enhancing the pharmaceutical industry will be, or partly is already, tangible along the whole value chain. Nevertheless, drug discovery and clinical research are the areas where the most spectacular milestones are expected. The global market for AI in drug discovery is forecasted to grow from some 1.5 billion U.S. dollars currently up to approximately 13 billion dollars by 2032. AI-based solutions in clinical research are projected to exceed seven billion U.S. dollars globally by the end of the decade. Global funding in artificial intelligence in drug discovery and design was highest in 2021, implying that the COVID-19 pandemic was the major catalyst for this development. In both, AI-enabled drug discovery and clinical trials, the U.S. and North America have a predominant role, maintaining a global market share of 50-60 percent.Speed, volume, cost, and precision
A view on the duration of biotech’s response to several health threats over the last two decades already shows how quick the development of the COVID-19 vaccine was. With the help of AI systems, Moderna’s vaccine was ready for clinical trials after just 42 days. Once a biological target is provided, AI systems can roam through reams of data in record time, identify patterns, and narrow down the number of drug or vaccine candidates. More recently, AI helped to find a new antibiotic. As antimicrobic resistance is one of the huge health challenges, this was a great success. AI efficiently screened 7,500 molecules, narrowing them down to 250 potential compounds in just 90 minutes. Potential time savings through AI usage for a new molecule until it becomes a preclinical candidate are between 40 and 50 percent, but some experts estimate that AI might cut down the time to develop drugs to one-tenth. Keeping in mind that 90 percent of clinical trial drugs never reach approval and that drug development costs can reach billions of dollars and last over a decade, AI imposes itself as (part of) the solution. Even more important than the above mentioned is that AI can help to tailor therapies more precise and personalized, for example through processing and interpreting huge amounts of genomic data sets.Completely “in-silico” is the next step
However, AI does not just mean sparing scientists and researchers boring and repetitive workloads along pre-programmed models. In 2022, the first drug completely discovered and designed by AI entered clinical trials and is currently in phase II. The developer, Hong Kong-based biotech company Insilico Medicine, uses Generative AI, the latest and far more creative generation of artificial intelligence. “In-silico” means that the drug has been developed fully ‘by computers’, referring to machine learning/artificial intelligence. While there have been other AI-developed drugs entering clinical trials, Insilico Medicine’s one marks the pioneering instance of a drug featuring both a unique AI-identified target and an original AI-crafted design. According to some projections, the global in-silico drug discovery market will reach six to seven billion U.S. dollars by the end of the decade.
