Mentioned in Video:
- 🧠 Deep Learning: This Disruptive Innovation Could Rewrite (or Erase) Us All: https://www.youtube.com/watch?v=J583J5Ts5uE
- ARK Invest's Newsletter Covering the AlphaFold Breakthrough (Issue 250): https://ark-invest.com/newsletters/issue-250/
- All In Pod Episode 41: Vaccine policy, Big Tech, DeepMind's latest breakthrough, wealth creation, opportunity & more: https://www.youtube.com/watch?v=FAtPReTswZo
- DeepMind's AlphaFold announcement: https://twitter.com/DeepMind/status/1418225170523705345
- DeepMind's website: https://deepmind.com/
- Central Dogma of Molecular Biology: https://atdbio.com/nucleic-acids-book/Transcription-Translation-and-Replication
- In giant leap for biology, DeepMind’s A.I. reveals secret building blocks of human life: https://fortune.com/2021/07/22/deepmind-alphafold-human-proteome-database-proteins/
- DeepMind’s AI predicts structures for a vast trove of proteins: https://www.nature.com/articles/d41586-021-02025-4
- AlphaFold predicts structures for nearly every protein in the human body: https://newatlas.com/biology/alphafold-structures-every-protein-human-body/
- Support the channel and get extra member-only benefits by joining us on Patreon: https://www.patreon.com/tickersymbolyou
🧬 #ARKG is a genomics-themed ETF managed by #CathieWood, the CEO and CIO of #ARKInvest. @ARK Invest also has many #genomics stocks in #ARKK, their flagship innovation fund. Several months ago, Cathie Wood spoke about the significance Alphabet's #DeepMind and their breakthrough with #AlphaFold. This past week, AlphaFold released and open-sourced their protein structure database, which offers the most complete and accurate picture of the human proteome. Many genomics #stocks will benefit from this incredible development, so we should what types of stocks just became some of the best stocks to buy now. Whether or not Google stock (GOOG) is one of the top stocks to buy now, I'm pretty sure their AlphaFold team will win a Nobel for this contribution.
Because of this breakthrough, we think that $100000 to $1 million is going to collapse to pennies. Provocative, it's one of the most provocative improvements we've ever seen in artificial intelligence. We didn't think we would see this for another six years.
Those were some highlights of Cathie Wood talking about the artificial intelligence breakthrough of DeepMind's Alpha Fold program. Last year. This past week, DeepMind opened up its Alpha Fold protein structured database for the world for free, which has massive implications for the future of health care and biology. It's not a stretch to say that the Alpha Fold team could win the Nobel Prize in Chemistry for this. So in this episode, I'll explain what you need to know as an investor, the types of new businesses this database could unlock and what this might mean for ARKG, ARK Invest's fund themed around the genomics revolution.
And of course, ARKK, their flagship innovation fund, which holds Cathie Wood's highest conviction stocks. If you enjoy this type of commentary and analysis, consider liking this video and subscribing to the channel with all notifications turned on. That way, you'll be the first to know when I come out with new research, regardless of how YouTube tunes its algorithm. Let's splice right into it. First, let's talk about what DeepMind and Alpha Fold even are. DeepMind is a company owned by Alphabet, the parent company of Google.
They combine machine learning, neuroscience, engineering, math, and powerful computers to solve different types of artificial intelligence problems. Their programs have learned to do things like diagnose eye diseases as effectively as the world's best doctors; save 30% of the energy used to keep data centers cool and predict the complex 3D shapes of proteins which could one day transform how drugs are invented and discovered. That last one is called Alpha Fold, which is the focus of this episode. Here's Cathie Wood explaining why Alpha Fold is so special in general, and its significance to ARK Invest specifically.
DeepMind, which is owned by Alphabet or Google, had a deep learning neural network work breakthrough called Alpha Fold. Alpha Fold is being applied to the genomic space and is helping researchers understand protein folding. Protein folding and just the structural shape of proteins are very important in terms of drug discovery. To give you a sense of what has just happened and how it's going to collapse the cost of a lot of research in health care and explode the knowledge base and the understanding of how the human genome works to discover the structure of a protein and understand it.
It took anywhere from a hundred thousand dollars to $1 million per protein structure. We now understand the structure of something like 150 to 200000 proteins, but there are roughly 170,000,000 proteins out there and that includes in plants and animals as well as human beings. Because of this breakthrough, we think that 100,000 to $1 million per protein structure is going to collapse to pennies. Provocative, it's one of the most provocative improvements we've ever seen in artificial intelligence. And the accuracy of these protein structures, as determined by Alpha Fold is more than 90%, a 70% improvement from two years ago.
We didn't think we would see this for another six years.
Okay, so Alpha Fold has just launched the protein mapping and drug discovery world about six years into the future in terms of accuracy and cost. But what even is protein mapping and how do we discover drugs? Here's a way to add some chemistry to your next date night. There's something called the Central dogma of molecular biology, which says that DNA makes RNA and RNA makes proteins. Wait. I don't think that's what Tinder meant by adding chemistry. Genetic material is stored in the form of DNA. In humans, the nucleus of our cells contains 3 billion base pairs of DNA distributed over 23 chromosomes. Altogether,
that stuff is called the human genome. The process of going from DNA to RNA is called transcription. RNA is what carries the information to make proteins. RNA is transported out of the nucleus of the cell and to the ribosome, where it directs protein synthesis. That process is called translation. The human genome contains around 30000 genes, each of which codes for one protein. Proteins are the building blocks of life and are required for the structure, function, and regulation of the body's cells, tissues, and organs, which means they have many important medical applications.
One example is insulin for diabetics. Another would be the antibodies that help develop vaccines. A third would be an enzyme that helps waste materials biodegrade faster. There are two important things that I think investors should really understand when it comes to proteomics or the study of proteins. First, the physical structure of a protein determines what it does. Hemoglobin, which is a protein found in red blood cells, has grooved pockets that bind to oxygen and transport it through our bodies. So seeing those grooved pockets in the structure of a protein gives people insight into what it binds to and why.
Second, proteins are pretty dang tough to model. Proteins are composed of hundreds of thousands of smaller chains called amino acids. Each amino acid has a unique size, shape, and electric charge, causing the protein to twist and bend in 3D space as these aminoacids fold into a stable position. This is the protein folding problem and to solve it, scientists have had to rely on tricky, time consuming and expensive experimental methods to build up a picture of the protein structure. That's that $100000 to $1 million price tag per protein structure that Cathie Wood was talking about.
Now that you're basically a doctor, let's talk about DeepMind's Alpha Fold program and their protein structured database. Alpha Fold is a software solution to this protein folding problem. You input DNA or RNA into Alpha Fold, and what you get out is a prediction of the protein structure without relying on these expensive physical and chemical experiments. DeepMind just published a database of the predicted structure of every protein in the human body. Before this, less than 20% of all human proteins had their structures fully defined by experimental methods.
So this is a huge leap forward. Like Nobel Prize in Chemistry huge. They also release the data for 20 other biologically significant organisms like lab mice and the malaria parasite. And they intend to add more protein structure predictions to the database until it covers all 130,000,000 genetic sequences that are known to encode the proteins for every living organism. That would be 700 times the number of the structures currently stored in the protein data bank, which contains all the protein structures verified by these expensive experiments so far.
And DeepMind is making this available to everyone in a searchable way for free. David Friedberg is the founder and CEO of the Production Board, or TPB, which is a San Francisco based investment foundry focused on improving technologies in agriculture, food, and the life sciences. Here he is on episode 41 of the All-in Pod with Jason Calacanis, David Sacks and, of course, Chamath Palihapitiya, going through a quick two minute example of the kinds of opportunities this database unlocks.
But let me just explain why this is relevant. The Delta variant, what it is, is the SARS CoV two RNA sequences, about 30,000 base pairs long. 10% of those are about 3000 base pairs make up the spike protein, which is the protein at the tip of the COVID virus, the coronavirus that gets into the cells. And for every ten people that are infected with coronavirus, there's about one nucleoside mutation, one of those base pairs changes, and the virus evolves. And we don't know how that change in that genetic code translates into a different structure of the protein.
And so we suddenly discover empirically by looking around, suddenly, all these people are getting more infected. Then we're getting infected before, we look at the genetic code and we're like, oh, here's the changes that happened. But we could have with this capability from Alpha Fold predicted. What changes make the spike protein do a better job binding to human's two receptors on the cells and getting into cells. And what other changes could be made in the whole genome of the SARS CoV two virus that could cause this virus to be more transmissible and more deadly.
All these sorts of factors because we can now estimate the physical form of that protein by changing the base pair. And so this tool that was released today, I think, highlights that over the next decade, these sorts of things that are going on with viruses, mutating and variants occurring that are affecting our population can be better estimated and tracked digitally. And it gives us the ability to start to prepare tools and defense mechanisms against them with new drugs and new variant models and new vaccines. Well ahead of the oh, my God,
we just got hit with a nuclear bomb. Let's clean up the mess in the future. So it's an exciting day, an exciting moment.
So with Alpha Fold, scientists can start predicting the most impactful genome mutations and their resulting effects on a protein structure. For a virus, this means they can analyze the whole genome to see what mutations would result in higher rates of transmission and fatality by seeing those changes in the structure itself. Knowing that in advance lets us prepare the exact targeted vaccines and treatments to stop those mutations from being effective before they can even come to pass. It could also help scientists quickly identify malfunctioning proteins and why they cause certain diseases already present in the human body and greatly accelerate the development of the drugs to treat them.
Certain companies are already in a prime position to make use of this unbelievable protein folding capability in the massive searchable database for drug discovery, protein engineering and synthetic biology. As a result, I think many companies inside ARKG and ARKK focused on bioinformatics, targeted therapeutics, molecular diagnostics, and agricultural biology are going to benefit greatly from this new capability. I wouldn't be surprised to see Alpha Fold or new technologies based on it get brought up in earnings calls and cause some of these companies to form entirely new business units.
Twist Biosciences, ticker symbol TWST, is a big name in synthetic biology. Their work focuses on their proprietary DNA synthesis platform, which works by manufacturing synthetic DNA on a Silicon chip. The applications for its platform range from drug discovery and development to digital data storage in DNA. Think about how powerful it is for them to be able to see the resulting protein structures of any DNA sequence and tweak it in software before starting their manufacturing process. Codexis, ticker symbol CDXS, is a protein engineering company that researches and develops enzymes and biotherapeutics for the pharmaceutical, food, industrial and life sciences industries.
Same thing, I expect them to benefit from Alpha Fold's database by being able to iterate on and optimize enzyme designs for their existing applications and discover new ones in their adjacent markets. In my opinion, the list of companies inside ARKG that could benefit from Alpha Fold is massive. Comment below with some of your favorite companies in this space and how they could benefit from this monumental capability and data set. Do you see any other types of companies benefiting as well? I'm excited to hear your thoughts.
Either way, I hope this episode helped you understand what DeepMind's Alpha Fold is, the specific kinds of companies that stand to benefit the most and why this is a big deal for all of humanity in general. If it did, let me know by investing in the like button and subscribing to the channel with all notifications turned on. That's a great way to invest in the channel that invests in you. Until next time. This is Ticker Symbol: You. My name is Alex reminding you that the best investment you can make is in you.
If you want to comment on this, please do so on the YouTube Video Here