The Royal Swedish Academy of Sciences has awarded this year’s Nobel Prize in Chemistry to three researchers whose work has fundamentally changed how scientists understand and engineer proteins. Demis Hassabis and John M. Jumper, the creators of the artificial‑intelligence system AlphaFold2, share half of the prize, while David Baker of the University of Washington receives the other half for his pioneering efforts in computational protein design.
Proteins, the molecular workhorses of living cells, adopt three‑dimensional shapes that dictate their biological roles. For decades, determining those shapes from a linear amino‑acid sequence required labor‑intensive techniques such as X‑ray crystallography, often taking months or years to resolve a single structure. The advent of powerful computers began to shift the field, but reliable prediction remained elusive.
Hassabis, a computer scientist and neuroscientist, and Jumper applied deep‑learning methods to the problem. Their AlphaFold2 system, released in 2020, demonstrated that a machine‑learning model could predict protein structures with a level of accuracy that many experts believed would not be achieved for another decade. Within months, AlphaFold2 generated predicted structures for nearly the entire human proteome, creating an open database now accessed by more than a million researchers worldwide. The achievement turned a long‑standing bottleneck into a rapid, publicly available pipeline.
In parallel, Baker pursued a different computational route. Beginning in the late 1990s with the development of the Rosetta software, his team combined physics‑based energy calculations with sampling algorithms to model protein folding. While Rosetta laid early groundwork, it could not consistently solve the most complex cases. Baker’s later work moved beyond prediction to design, enabling scientists to specify a desired function and then generate a protein sequence likely to perform that role. In 2022 his group introduced “protein hallucination,” a neural‑network technique that imagines entirely new protein shapes not found in nature, expanding the toolbox for creating bespoke enzymes, sensors, and therapeutic candidates.
The Nobel citation links the three laureates under the broad theme of protein research, highlighting both the predictive power of AlphaFold2 and the generative capacity of Baker’s design platforms. The prize underscores a broader shift in chemistry: computational methods are no longer auxiliary but central to discovery. Hassabis’s background in AI and his company DeepMind—acquired by Google in 2014—illustrates how expertise from computer science now drives breakthroughs in a traditionally experimental discipline.
Practical implications of these advances are already evident. Accurate structure prediction aids the study of diseases linked to protein misfolding, such as Alzheimer’s and Parkinson’s, by revealing how aberrant shapes arise and suggesting points of therapeutic intervention. Baker’s designed proteins hold promise for creating novel enzymes that could degrade plastics or serve as next‑generation vaccine components, potentially improving immune responses.
Nevertheless, the laureates and the Nobel committee acknowledge that challenges remain. Predicting a protein’s static structure does not automatically explain its behavior within the complex environment of a living cell, and designing a sequence on a computer does not guarantee correct folding in vivo. The awards mark a milestone rather than a final solution, signalling the beginning of an era where computation is a primary instrument in chemistry.
With these new tools now in the hands of the global scientific community, the next chapter of protein research will be written by those who apply, refine, and expand upon the methods introduced by Hassabis, Jumper, and Baker.
