|
Below is an excerpt from a complimentary research note written by our Healthcare Team of Tom Tobin, William McMahon, and Justin Venneri. We are pleased to announce our new Sector Pro Product Health Care Pro. Click HERE to learn more. |
‘It will change everything’: DeepMind’s AI makes gigantic leap in solving protein structures - The Protein Folding Problem solution announced by DeepMind’s AI (Google) won’t get the kind of attention that the Human Genome Project did, but it should, and likely more. The pace of AI solutions for unsolvable problems is bleeding into the backdrop of NBA scores and celebrity scandal.
“Time travel? Cool.”
The reason THIS solution is so exciting (or frightening) is what it will unlock in terms of our understanding of human biology at the systems level. It’s a useful thing to know what your DNA sequence is, but understanding what it means has so far been a game of indirect evidence.
A good example of a gene-based experiment is one that compares normal to diseased. The experiment design is “consider two cells, such as a normal liver cell and a cancerous one, how do their genes differ?"
There are safe to say millions of steps between a mutation in a gene and cancer. In a gene, there is information written in a linear sequence, and a protein takes that sequence, builds a string of amino acids, and then “folds” it into a 3 dimensional shape, not just a random coil, but a finely tuned shape with an incredibly specific function that interlocks with tens of thousands of other proteins.
Bridging the gap from a gene to functioning protein cracks open the door to systems biology where the research can see through from a mutation in a gene all the way to the impact on a whole person.
It's incredible stuff where the pace of “what if” questions and their answers should accelerate. It’s a world where a researcher can type up the genetic code and see the output at the cell level, tissue, or whole organism. It may not be tomorrow, but this is the advent of designer biology.
For our stocks, we would expect the demand to sequence DNA and RNA will accelerate and provide a tailwind to ILMN. We don’t have a good way to “play” protein folding yet unless there is a way to get long and bioinformatics computing power and talent which will be in high demand.
