Bio2040 - Bottlenecks & Future of Science, Healthcare & Drug Discovery

One of the most exciting conversations I have had all year. Abhi, co-founder and CEO of roots is is a software engineer turned farmer who is really pushing LLMs to the edge of their abilities. In this interview, we discover how he has been using Google Gemini's large context windows to feed it thousands of agriculture science papers and discover insights relevant for his agriculture startup Living Roots. His insights have already generated 30% yield increases for smallhold farmers in the first season, at very limited cost!At roots, Abhi and the team are even going so far as to combine Reinforcement Learning with LLMs reasoning abilities to predict the outcome of agricultural interventions.During this interview, I really got to think that every research lab on the planet should have the pipeline that Abhi has built. Please watch for a really insightful and fun conversationCheck out rootshttps://roots.ag/Follow Abhi on LinkedInhttps://www.linkedin.com/in/abhimagarwal/Follow Flavio on LinkedIn for more great content on the intersection of AI and Agriculture/Climatehttps://www.linkedin.com/in/flaviorump/Medium: https://medium.com/@flaviorumpTwitter: https://twitter.com/flaviorump

In Bio2040's latest episode we had the chance to talk to Sam Lee, a long-time veteran of the biotech space. This episode covers some of the history and fundamentals of the synthetic biology and biomanufacturing space.

What are the main challenges in translating scientific innovation from the bench-side to the bedside? Physician and long-term Pharma executive Nadine Martin is now in charge of innovation management at sitem-insel AG. A translational unit of the Inselspital in Bern.

SEED2019 is a scrappy conference targeting innovation within the science publishing & funding process. It will bring together a diverse set of stakeholders and be followed by an incubator to implement the best ideas. We discuss problems and ideas on how to best solve the most pressing problems of academia.

How Open Source Drug Discovery can lower the risk and cost of drug discovery. M4K Pharma is pioneering a completely new model, where patents are not filed but data is shared. Yet they are still able to make money.

Peer review is opaque and a large frustration for many scientists. Preprints are becoming more prevalent, but we still need to assess them critically. We shed light on some of the problems and solutions to this with preprint peer review clubs. PREreview is a platform to record these journal club conversations so that the rest of the world can benefit from them too.

Oskari and Flavio talk about how robots can help humans in biomedical research. UniteLabs is building lab automation systems to automate boring tasks so the highly trained researchers can focus on more high-level activities. We also talk about other bottlenecks in biomedical research and revisit some old friends from previous podcast guests.

What does Open Research mean for the Gates Foundation? Why does it matter? How are they implementing it? We speak with Ashley Farley who is in charge of those initiatives at the Gates Foundation.

In the 2nd part of the interview with Prof. Ernst Hafen we learn about how the birth of the Bio-Technopark Schlieren shows that if innovative entrepreneurs come together with great scientists, magic can happen. We also learn about a big bottleneck in translating science from academia into startups and how the future of medicine is personal.

Lenny is the founder and CEO of protocols.io, where scientists can freely share scientific protocols with each other. Similarly to github, others can then try them out in their own lab, comment on them, improve them or fork them. This leads to many benefits for both the original author as well as the scientists who get access to new protocols. 11,000 scientists are already signed up and more are joining each day.

In this episode, Prof. Ernst Hafen from ETH Zurich tells us how citizens can play a very important role in biomedical research and drug discovery. Recruiting the right patients for a study is a big challenge for biotech and pharma companies. He presents us his radical initiative midata.coop, which allows patients to store their health data in a secure way and allows them to participate in clinical studies. Part 1 of a very interesting interview.

In this episode, we learn how Huntington's disease researcher Rachel Harding uses an open lab notebook to share her findings with the scientific community and writes a blog for a wider audience. Most science is not conducted like that and findings take a long while to be disseminated. What are the main benefits of doing open research? Why haven't more researchers adopted this yet? What areas of biomedical science are most suitable for open research?

Sarah Constantin is a mathematician & data scientist with a PhD from Yale. We discover how computer vision and machine learning can be used to do drug discovery what the biggest challenges in today's cancer research with potential solutions

Prof Paul Glasziou is a practitioner and widely published clinical research expert. He explains how we waste billions of dollars on biomedical research. Over 50% of all health research does not get published at all. From the published research, 50% is not usable in practice because things are missing. And from the 25% which is left, half contains design flaws. Hence, over 87.5% (which they rounded down to 85%) of health research is avoidably wasted according to Glasziou. We explore how this waste happens and what can be done about it.

In this podcast, we get to know Marcio van Muehlen, a biological engineer from MIT turned entrepreneur & product manager in Silicon Valley. He was working on cancer diagnostics through blood sampling and wanted to have better access to the work from other scientists in the field. There he recognized early on that we need a github for science and published a widely referenced article on it in 2011. We talk about why we need a github for science, what has happened since he wrote the article and what is still preventing us from having more open collaboration.