An obvious near future implementation of this work is in drug development pipelines. By modeling out a limited number of likely proteins to test for efficacy, the cost and efficiency in bringing new drugs to animal/cell models, clinical trials and eventually marketplace for routine use, can be reduced tremendously. The historic models have been to do trial and error testing on hundreds if not thousands of potential protein iterations of potential therapeutics, which costs huge amounts and takes years of development. Limiting the number of proteins to be tested, by modeling the most likely to be successful, these cost and time savings could be realized and change the entire face of drug development from its current paradigm. Perhaps, paired with legislative incentives, these savings could make their way into savings for consumers?
CTO | AI for Cancer | Applied AI | Board Member | Customer Engineering | Developer Relations | SWE | Entrepreneur
Bravo, AlphaFold team! Lee Hood remarked to me in 2020 that applying “hyperscale AI” to “phenomics and cancer” is Nobel level work. Then this. Wow 🤯 a huge congrats to the team. The first Nobel! Lee’s convinced that once the world assembles sufficient data, we will see healthcare shift from sick care to wellness prevention. Medicine could forever change, “and you won’t see a doctor without being sequenced, or have your epigenome sequenced 1-2 times a year.” Lee and several of the top minds in this space, as well as nascent startups, will convene on October 30th in Boston at our first AI Summit for Cancer. Including deep mind. A “cure” to more cancers may lurk among them, as we have seen with HER2 positive cancer trials today. My Mom would be so encouraged, may she rest in peace. My brother David is coming to the event too. Can’t wait! Again, what a terrific day for AI in healthcare and life sciences. 🎉