Bizmarketeer - Webinars & Events

Webinars & Events 8 Min Read

Tesseract Talks – Cresset

In this episode of Tesseract Talks, host John K. Thompson, data and AI thought leader and author of The Path to AGI, speaks with Dr. Mark Mackey, Chief Scientific Officer at Cresset, who has spent over two decades building computational tools for pharmaceutical R&D, to explore how AI and computational chemistry are reshaping drug discovery.

The conversation centers on one of the most complex and underappreciated aspects of drug development: ensuring that a molecule not only works biologically but can actually function as a viable drug inside the human body. Mark introduces the concept of ADME (absorption, distribution, metabolism, excretion), explaining that success in drug discovery goes far beyond finding a molecule that binds to a target protein. The real challenge lies in predicting whether that molecule survives digestion, reaches the bloodstream, avoids breakdown, enters cells, and performs its intended function without harmful side effects.

From there, the discussion shifts to the limitations of traditional computational approaches and how modern AI, particularly machine learning and deep learning, is improving predictive accuracy. Rather than replacing scientists, AI acts as a powerful pattern-recognition engine, helping researchers decide which molecules are worth synthesizing and testing. Even modest improvements in prediction accuracy can significantly reduce the number of experiments needed, lowering costs and accelerating timelines.

Mark also highlights the growing industrialization of drug discovery. What was once a slow process is evolving into a more automated pipeline, combining generative AI for molecule design, predictive models for screening, and robotics for synthesis. This shift mirrors the transition from handcrafted production to scalable manufacturing, enabling faster iteration cycles and more efficient experimentation.

The episode closes with a grounded perspective on AI’s role in life sciences. While AI is already delivering value in optimization, automation, and cost reduction, it is not a silver bullet. The biggest bottleneck remains our incomplete understanding of human biology, which is the primary reason why most drugs fail in clinical trials. The real opportunity lies in using AI to make failure cheaper and faster, rather than eliminating it entirely.

Takeaways

  • Drug discovery is fundamentally constrained by ADME challenges, ensuring a compound can survive, travel through, and function within the human body, not just bind to a target.
  • AI excels at pattern recognition in complex biological data, enabling better predictions of drug properties and helping researchers prioritize which compounds to develop.
  • Even imperfect AI models create value, as being “better than a coin toss” can significantly accelerate discovery and reduce wasted experimentation.
  • Generative AI plays a supporting role, guiding molecule design when combined with strong predictive models, rather than solving drug discovery independently.
  • The industry is shifting toward an industrialized model, integrating AI-driven design,
    predictive modeling, and robotic synthesis into scalable pipelines.
  • The biggest limitation is not computational power but biological understanding,
    which is why high failure rates in clinical trials will persist in the near term.
  • The most effective use of AI is automation, freeing scientists from repetitive tasks
    and allowing them to focus on high-value insights and innovation.
  • Future leaders in AI-driven drug discovery will be those who integrate AI pragmatically
    into workflows, balancing real value with realistic expectations rather than hype.

Please complete the form below to access the report

Knowledge Depot (part of Communitize Ltd) would like to contact you with details of other services we provide. If you consent to us contacting you for this purpose please tick to say how you would like us to contact you.

We would like to pass your details onto Software Advice Inc, providers of this content. Software Advice may wish to contact you via email or telephone with details of products and services that they provide.