Nobel in Chemistry: AI's Protein Breakthrough

⏱️ 13 min read     

Today, we are designing life-saving proteins on demand, reshaping medicine as we know it—thanks to artificial intelligence. The Nobel Prize in Chemistry that was awarded in October of 2024 celebrates just that! 

The pioneering work of David Baker, Demis Hassabis, and John Jumper, harnessed AI to predict and create proteins, a breakthrough that could transform everything from drug discovery to bioengineering. Their award was the second Nobel that same month, spotlighting AI. Is this the dawn of a new era in science? Well, it's time to find out.

Welcome to That's Life, I Swear. This podcast is about life's happenings in this world that conjure up such words as intriguing, frightening, life-changing, inspiring, and more. I'm Rick Barron your host. 

That said, here's the rest of this story  

On Wednesday, October 9, 2024, the distinguished Nobel Prize in Chemistry was awarded to three upcoming pioneering scientists. Their groundbreaking work has unlocked the remarkable potential of advanced technologies, including artificial intelligence, to unravel the mysteries of protein structures - the fundamental building blocks of life.

2025 Nobel Prize Chemistry Winners: left to right, David Baker, Demis Hassabis and John Jumper. Courtesy of Institute of Protein Design

The esteemed three laureates are Demis Hassabis and John Jumper of Google DeepMind, who leveraged the power of AI to predict the intricate shapes of millions of proteins. Alongside them, David Baker of the University of Washington, was recognized for his innovative use of computer software to invent entirely new protein structures.

As Johan Aqvist, a member of the Nobel Committee for Chemistry, stated, the impact of this year's prize-winning achievements was "truly huge." He emphasized that understanding the form and function of proteins is essential to comprehending the fundamental workings of living organisms. The work of these visionary scientists has profoundly advanced this critical field of study.

The 2024 Nobel Laureates in Chemistry have pushed the boundaries of scientific knowledge and ushered in a new era of technological skill to uncover life's most intricate secrets. Their pioneering efforts have left an indelible mark on the scientific community and hold the promise of transformative breakthroughs for future generations.

The process of predicting the shape of proteins once took months if not decades. However, revolutionary AI models like AlphaFold , a protein structure database, have made it possible to accomplish this task in hours or even minutes!

This increased speed has real-world applications. AlphaFold has been extensively cited in over 20,000 scientific studies, enabling biochemists to accelerate the discovery of new medicines. As Dr. Jumper stated, "We can draw a direct line from our work to improving people's health."

Furthermore, these advancements could lead to the development of novel biological tools, such as enzymes capable of efficiently breaking down plastic bottles and converting them into easily reusable and recyclable materials.

The recognition of the 2024 Nobel Prize in Chemistry was the second Nobel awarded that involved artificial intelligence, underscoring the growing significance of this transformative technology in scientific research.

AI is revolutionizing scientific research. Frances Arnold, 2018 Nobel Laureate, stated: "It is supercharging our ability to explore previously intractable problems."

The selection of these Chemistry Laureates signals a shift towards more computational study, which American Chemical Society President Mary Carroll says "may improve efficiency while reducing lab work."

However, AI also introduces risks, as DeepMind's Demis Hassabis cautions: and I quote, "It has extraordinary potential for good, but also can be used for harm." End quote.  There are concerns that AI could enable the creation of new bioweapons. Nobel Laureate David Baker was among 90+ scientists seeking to regulate bioweapons manufacturing equipment.

The Nobel Committee's choice reflects AI's growing influence in transforming scientific discovery and the dual-use challenges it presents.

Proteins and enzymes are the microscopic mechanisms driving all life. Their three-dimensional shapes define their functions. For decades, scientists struggled to solve the "protein folding problem" and determine individual protein structures.

So, the question comes up. What is protein folding? Let me provide an analogy.
 
Protein folding is like origami for molecules. Just as a flat sheet of paper transforms into a specific, intricate shape when folded along precise lines, a protein starts as a long chain of amino acids and folds into a unique three-dimensional structure. The final shape determines its function, much like how a folded paper crane or flower reveals its purpose and form. But unlike origami, proteins must fold flawlessly to work properly—one wrong fold, and the result could lead to malfunction or disease.

Demis Hassabis, a Cambridge-educated computer scientist and neuroscientist, co-founded DeepMind in 2010. Google later acquired DeepMind for $650 million. DeepMind aimed to build artificial general intelligence and solve key scientific challenges, including protein structure prediction. As a teenager, Demis was the second-highest-ranked chess player under 14 in the world, and he began designing video games professionally before attending college.

AlphaFold leverages neural networks to predict the 3D structures of proteins. These intricate shapes determine how molecules bind to and interact with proteins - a crucial factor in drug development. Neural networks enable computers to process massive amounts of data, allowing them to learn and perform tasks that were previously out of reach. These systems power technologies like facial recognition, voice recognition, and online chatbots.

John Jumper, the youngest Chemistry Nobel laureate in over 70 years; co-led AlphaFold's development at DeepMind. After earning degrees from Vanderbilt, Cambridge, and the University of Chicago, Jumper joined DeepMind in 2017.

In 2018, Jumper's team entered AlphaFold in the Critical Assessment of Structure Prediction competition, outperforming competitors in solving the longstanding "protein folding problem."

In 2020, the AlphaFold2 update from Google researchers cracked the protein folding problem, predicting shapes with accuracy rivaling physical experiments.

"We just woke up that day and knew: this is a different biology," remarked Columbia University's Mohammed AlQuraishi.

With AlphaFold2, the team calculated the structures of all human proteins and eventually predicted 200 million known protein structures.

DeepMind founder Demis Hassabis modeled the company after Bell Labs, a prolific R&D powerhouse. He noted that cutting-edge AI research requires substantial resources and computing power, which the private sector can help fund.

David Baker, a Seattle native, pioneered new protein design before the latest AI breakthroughs. In 2003, Baker's team created the first entirely new protein, Top7, using computer modeling to find an amino acid sequence for their desired structure.

Baker recalled the "amazing moment" when the bacteria-produced protein matched his computer model - "opening up a completely new world of protein structures."

The researchers began by envisioning the desired protein shape and utilized a computational tool called Rosetta. This software analyzes existing protein databases to identify amino acid sequences likely to produce the targeted structure.

Baker described the "incredible moment" when a protein engineered in bacteria from the suggested amino acid sequence closely matched the predicted structure from the model.

"This breakthrough unlocked an entirely new realm of protein structures that had previously been unknown," said Dr. Aqvist of the Nobel committee.

Dr. Baker realized he could create sophisticated proteins to tackle challenges like Alzheimer's. His lab combines neural networks and the Rosetta software to predict and generate novel protein structures.

These laboratory-designed proteins have already formed the basis for potential medical treatments, including a Covid-19 nasal spray and a celiac disease medication. One of Dr. Baker’s protein designs, SKYCovione, was approved as a COVID-19 vaccine in South Korea in 2022.

In addition to his academic work, Dr. Baker has co-founded over 20 biotech companies. When asked about a favorite protein, he replied: "I love all proteins. I don't want to pick favorites."

When the Nobel committee called David, Demis, and John to notify them that they had won, there were some amusing moments:

• David Baker was sleeping when they reached him. His wife started screaming so loudly he could barely hear the announcement.
• John Jumper had resigned himself to not winning that year, telling his wife, "I guess it's not this year," just 30 minutes before the news conference.
• After receiving the call, Jumper recorded a video sharing the news with celebrating colleagues on a video call.
• Demis Hassabis said the news "hadn't sunk in" and that his typical work plans for the day would have to change.

The Nobel laureates' reactions ranged from sleep-deprived surprise to joyful disbelief, providing some lighthearted moments amidst the prestigious recognition of their groundbreaking scientific achievements.

What can we learn from this story? What's the takeaway?

·       Accelerating Discovery: A.I. dramatically shortens the time needed to develop new proteins compared to traditional trial-and-error methods. This can speed up the creation of novel treatments for diseases like cancer, Alzheimer’s, and infectious diseases.

·       Precision Medicine: A.I. can design proteins tailored to individual needs, paving the way for personalized therapies that target diseases at their molecular roots with minimal side effects.

·       Expanding Possibilities: A.I. opens doors to creating proteins that don’t naturally exist, enabling us to tackle previously "undruggable" conditions or engineer entirely new biological functions.

·       Global Health Impact: By making protein design more efficient and scalable, A.I. can help address challenges like antibiotic resistance, pandemics, and vaccine development, reaching populations with limited healthcare access.

·       Cross-Disciplinary Innovation: The integration of A.I., biology, and medicine demonstrates the power of collaborative approaches to solving complex problems, inspiring similar innovations in other fields.

In the end, A.I. in protein design exemplifies how technology can revolutionize science and medicine, transforming our ability to heal and protect human life. It reinforces the importance of investing in interdisciplinary research and ethical oversight to harness these breakthroughs responsibly and equitably.

Well, there you go, my friends; that's life, I swear

For further information regarding the material covered in this episode, I invite you to visit my website, which you can find on Apple Podcasts for show notes and the episode transcript.

As always, I thank you for the privilege of you listening and your interest. 

Be sure to subscribe here or wherever you get your podcast so you don't miss an episode. 

See you soon.