Protein Folding and AI (Nobel Prize 2024) - Periodic Table of Videos

Introduction

This year's Nobel Prize for Chemistry has generated excitement, as it delves into the realm of protein folding — a topic that, while not traditionally rooted in chemistry, is critical for numerous scientific advancements, including understanding life and curing diseases. The prestigious award was shared by three laureates: American scientist David Baker from the University of Washington, and two computer scientists, Demis Hassabis and John Jumper from Google DeepMind. Their work centers around the intricate structures of proteins, biological molecules essential to all organisms and encoded by DNA.

Proteins are composed of chains of amino acids, sometimes numbering in the hundreds, which fold into specific shapes. The complexity of these molecules leads to an astronomically large number of potential structures—approaching or even exceeding the number of atoms in the universe. The challenge lies in deciphering why proteins adopt particular shapes. The prize acknowledges two distinct approaches to this overarching problem.

David Baker has utilized advanced deep learning software to design proteins that fold into predetermined shapes, focusing primarily on shorter amino acid chains. Conversely, the DeepMind team has worked from the opposite direction, employing artificial intelligence to predict the shapes that longer amino acid sequences will adopt. An annual competition started in 1994 has aided this effort, challenging computer scientists to predict the structures of 100 proteins, whose configurations have already been accurately determined through experimental methods like x-ray crystallography.

Understanding the three-dimensional structure of proteins is crucial. While scientists can determine a protein’s sequence, identifying its shape is essential for understanding its function. In a human body, there are millions—perhaps billions—of different proteins, and knowledge of their sequences alone is insufficient without insight into their structure.

Despite the debate around the fit of AI-influenced solutions within the realm of traditional chemistry, the fundamental importance of the problem justifies the recognition of this work. DeepMind's AlphaFold software achieved a breakthrough in 2020 when it accurately predicted protein structures with 90% accuracy for individual atom placements. This advancement has drastically increased the number of understandings of protein structures, boosting the known figures from 200,000 to around 200 million—an endeavor that previously required years of detailed laboratory work.

David Baker's contributions also include designing a novel protein composed of 93 amino acids that closely matched the predicted structure, showcasing the potential for new applications in various fields.

Both Baker and the DeepMind teams represent larger groups of dedicated scientists collaborating to tackle this complex challenge. Hassabis has expressed ambitions not only to challenge intelligence but to leverage it to solve various scientific problems, including weather prediction and game strategy. This Nobel recognition underlines how innovative computer approaches can offer solutions to longstanding scientific dilemmas.

---

Keywords

• Protein folding
• Nobel Prize
• David Baker
• Demis Hassabis
• John Jumper
• AI
• Deep learning
• AlphaFold
• Amino acids
• Protein structure

FAQ

Q: What is the significance of the Nobel Prize awarded for protein folding?
A: The Nobel Prize highlights the advancements in understanding protein structures, which are essential for unraveling biological processes and developing new treatments for diseases.

Q: Who were the recipients of the Nobel Prize in Chemistry for 2024?
A: The prize was awarded to David Baker, and the DeepMind scientists Demis Hassabis and John Jumper.

Q: How does AI relate to protein folding?
A: AI, particularly through DeepMind's AlphaFold, has revolutionized the prediction of protein structures, allowing for quicker and more accurate models than traditional experimental methods.

Q: Why is the structure of proteins important?
A: Protein structure is crucial in determining their function within biological systems. Without structure, understanding the role a protein plays is nearly impossible.

Q: How has the understanding of protein structures progressed recently?
A: The number of comprehended protein structures has surged from 200,000 to 200 million, thanks to advances in AI and deep learning techniques.