Pharma R&D Today
Ideas and Insight supporting all stages of Drug Discovery & Development
Disrupting innovation: From geography, penguins and ocean currents to diagnosing disease by its smell
Posted on November 15th, 2017 by Dr Andrew A. Parsons in Pharma R&D
How to become aware of what I don’t know? These insights can stimulate disruption. They help us see a whole new world or perhaps a pattern of activities that need to change. As Donald Rumsfeld described, it is those things we don’t know we don’t know that can be issues for the future. They can also be lost opportunities in scientific understanding and commercial success.
With so much information being circulated in the world, how can we find ways to link information to areas we would perhaps not normally be aware off? There are huge opportunities for artificial intelligence in this area, and a gold test would be the ability to find something surprising and unusual that is useful. I remember the days of libraries, where reading something different could spark an idea. Developing this depth of expertise and a breadth of knowledge in associated areas is thought to enable innovation.
As an academic researcher, I was proud to have been awarded an Alexander von Humboldt Fellowship to work in Munich. This was an amazing experience in terms of professional and personal development as it provided a chance to develop in-depth scientific knowledge and a breadth of contacts within a network of research excellence in physics, the natural sciences and the humanities. This allowed unusual and useful conversations to take place, which really reflects the career of von Humboldt. He developed a systems concept of the Cosmos, was one of the first to suggest that South America and Africa had been joined together, and lends his name to Penguins and Ocean Currents. The Foundation is still supporting innovation today. It was in a recent article the Institutes magazine where I learned of the developments of unique “smells” that are linked to disease (1).
I had known that dogs can be trained to “smell” disease, but the small report describing the varied research funded by the Humboldt Foundation provided a link to how this could be used in R&D. A recent article highlights the diagnosis and classification of 17 diseases by pattern analysis of exhaled volatile organic compounds (2). The use of a gold-based nano-array biosensors coupled with artificial intelligence has provided the impetus for these recent developments (2,3). As they point out, although it is not common in modern practice, physicians have been using odor to check the health of their clients for thousands of years. This technology is likely to provide data that will bring smell back to the table, at least to support diagnosis in the first instance.
It might be interesting to consider how the metabolism of small molecules can impact the pattern of exhaled volatile organic compounds. This might develop specific PK/PD relationships for specific molecules under investigation. It may also support the development of highly cost-effective outcome measures for developing personalized medicine regimens for treating cancer and other diseases. Developing a simple companion diagnostic to measure drug action is a vision of many research projects to aid patient selection. There are highly effective imaging approaches that can be used in these areas, but these involve highly technical equipment, trained professionals and provide a complex patient journey with multiple hospital visits. Ultimately, biosensor technology has many potential technical and economic benefits.
What was interesting to me in this recent experience was the positioning of several topics that triggered some interest, understanding and ideas. This approach has been a cornerstone of the intention of the von Humboldt and other learned societies over many years. The challenge in our data-rich world is how to assemble this breadth and depth and suggest opportunities for follow-up? This generates a critical interface with the user which will determine future action. To really create disruption, AI will deliver the right communication to impact behavior, investment and energy. Developing this ability to spot patterns and trends that creates a change in human behavior is critical.
The challenge for AI companies to connect the dots to provide surprising, unusual and useful insights is significant. However, the race is on. There are several companies that already show the potential to disrupt the research and development cycle. Hopefully there will many more instances of findings that drive action. The problem then will be, which one to follow? However, that is always a good problem to have and no doubt will ultimately change the way we operate. One key area will be how we create a competitive pipeline, while a second is perhaps from a pharmacovigilance perspective: AI will enable effective and rapid monitoring of the real-world experience and safe use of new products.
- Humboldt Kosmos, 2017. Mr Haick, what does the breadth tell us about our health. 107: p11.
- Nakhleh et al., 2017. Diagnosis and classification of 17 diseases from 1404 subjects via pattern analysis of exhaled molecules. ACS nano 11: 112-125.
- Peng et al. 2009. Diagnosing lung cancer in exhaled breath using gold nanoparticles. Nature Nanotechnology 4: 669-673.
All opinions shared in this post are the author’s own.
R&D Solutions for Pharma & Life SciencesWe're happy to discuss your needs and show you how Elsevier's Solution can help.
Dr Andrew A. Parsons
Director of Reciprocal Minds Limited & Chairman of Pharmasum Therapeutics AS
- Emulating start-up success with purpose-built data tools
- Finding Answers Where You Least Expect Them
- Machine Learning Could Make Drug Discovery Faster, Cheaper, Better
- AI Can Work Wonders, But Can It Create Miracle Cures?
- The Future of Chemistry R&D Will Depend on Deep Learning