Generative AI Developed Drug Reached Clinical Trails Phase 2

Healthcare stands to benefit greatly from advanced AI applications, and companies like Insilico Medicine are at the forefront of using AI to develop drugs more efficiently. Recently, Insilico Medicine announced that it has successfully entered phase 2 clinical trials with the first AI-developed drug. Alex Zhavoronkov, the co-founder of Insilico Medicine, discussed the significance of this achievement on the London Futurists Podcast.

The Focus: Idiopathic Pulmonary Fibrosis (IPF)

The AI-developed drug targets Idiopathic Pulmonary Fibrosis (IPF), a condition characterized by the thickening and scarring of lung tissue. IPF primarily affects individuals over the age of 60 and can be fatal. Unlike other AI drug development companies that typically focus on well-known proteins, Insilico Medicine has identified a new target protein for IPF. By analyzing extensive datasets, Insilico’s AI narrowed down a list of potential proteins to 20, eventually identifying a favored candidate. They then utilized generative AI models, a type of AI known for its ability to create and criticize content, to identify molecules that could disrupt the target protein.

Generative AI and Drug Development

Generative Adversarial Networks (GANs) and Generative Pre-trained Transformers (GPTs) are two types of generative AI models. Insilico Medicine employs up to 500 different generative AI models and provides them with the protein’s crystalline structure to design molecules that effectively bind with the target protein. Over a few days, these models compete to find the most suitable molecule for the job. Human chemists from various Contract Research Organizations (CROs) review the most promising molecules, and around 15-20 are synthesized and tested. The best-performing molecules’ characteristics are then fed back into the generative AI systems for further refinement.

Progression to Clinical Trials

The resulting molecules underwent testing for efficacy and safety in animals such as mice and dogs. In 2021, Insilico Medicine moved on to phase zero of the clinical trial process, conducting a preliminary safety test on eight healthy volunteers in Australia. This was followed by a phase one clinical trial, involving a large-scale safety assessment on healthy volunteers in New Zealand and China. Due to IPF being a chronic condition, the phase one trial required thorough testing, as patients would be taking the drug for an extended period. Currently, Insilico is preparing for the phase two study, where patients with IPF will receive the drug in China and the USA. Recruiting a sufficient number of patients with a good life expectancy remains a challenge.

Reducing Failures and Consolidation

According to Zhavoronkov, Insilico Medicine has shortened the typical six-year drug discovery and development process by a couple of years. The most significant benefit of AI in drug development lies in reducing the failure rate of candidate molecules, as approximately 99% of them fail. The AI-driven approach offers improvements in this regard. In recent years, the AI drug development sector has experienced consolidation, with fewer companies remaining after the initial hype during the COVID-19 pandemic. Surviving companies, including established ones like Schrödinger Inc., are now better positioned to identify genuine opportunities.

New Technologies and Future Prospects

Companies that weather the consolidation process have numerous opportunities ahead. Quantum computing, in particular, holds promise, with Zhavoronkov anticipating significant impacts within the next five years, possibly even within two years. Insilico Medicine utilizes IBM’s 50-qubit machines for quantum computing and acknowledges IBM’s cautious approach following the over-hyping of its AI suite, Watson. Other tech giants like Microsoft and Google also have ambitious plans for quantum computing. Generative AI for drug development could be among the first highly valuable use cases for this technology.

While AI-led drug discovery remains slow and costly compared to traditional methods, advancements like GPTs provide hope for a better understanding of human biology. While pharmaceuticals alone may not cure aging in the near future, improved health outcomes and longevity may be achievable if individuals maintain good health during middle age, thanks to advancements in today’s technologies.

News Source