2024 Nobel Prize in Chemistry(1)
Reason for Award
Development of computational methods for protein design
Laureates
United States of America
Explanation
Proteins, the building blocks of our bodies, are strings of 20 different amino-acid “beads” that fold like origami. Dr. Baker used computers to design brand-new proteins that do not exist in nature. It is similar to deciding the shape of a LEGO model first and then letting a computer tell you which bricks you need and where to place them. The designed proteins can clean dirty water, break harmful chemicals, or become medicines. His work makes it easier and greener to create helpful products for people and the planet.
Related Keywords
de novo design
A methodology that starts from a desired 3-D fold or function and computes an amino-acid sequence from scratch instead of modifying existing proteins. It combines energy minimization with statistical patterns to generate folds absent in nature. Applications span therapeutics and industrial enzymes. Once computationally prohibitive, de novo design has been accelerated by cloud, GPU, and AI resources. Growing numbers of successes are revealing generalizable design principles.
Rosetta
A protein structure prediction and design suite developed at the University of Washington. It samples folding pathways via fragment reconstruction and stochastic search, evaluated by empirical energy functions. Modular scripting allows diverse tasks such as enzyme active-site design and antibody optimization. Distributed as open source, it is continuously extended by a global community. Volunteer computing through Rosetta@Home supplies massive computational power.
energy landscape
A high-dimensional space where each conformation of a protein has an associated free energy. Natural proteins display a funnel-shaped landscape whose global minimum represents the native fold. Design seeks to engineer such a landscape by creating a deep well around the target structure. Coarse potentials can allow misfolds, so precise terms for solvation and electrostatics are added. AI models trained on experimental data now refine energy landscape evaluation.
self-assembling nanocage
Multiple designed protein subunits assemble with defined symmetry to create nano-sized hollow cages. They are attractive for vaccine antigen display and drug-delivery carriers. Design controls hydrophobic packing and hydrogen-bond networks at interfaces to achieve high assembly specificity. Cryo-EM now resolves their structures near atomic resolution, guiding functional modifications. Stimuli-responsive nanocages that open or close with temperature or pH are being developed as smart materials.
artificial enzyme
A computationally designed protein that catalyzes chemical reactions absent from nature. Quantum-chemical models define the reaction mechanism, and Rosetta optimizes active-site residue placement. Successful cases include enzymes accelerating Diels–Alder cycloadditions and aromatic defluorination. After iterative optimization, catalytic efficiencies approach those of natural enzymes. Such biocatalysts promise greener industrial chemistry.
ProteinMPNN
A transformer-based sequence generator that proposes optimal amino-acid sequences for a fixed backbone in seconds. It is orders of magnitude faster and more accurate than traditional packing calculations. Statistical patterns learned from large datasets complement energy evaluations, raising experimental success rates. High sequence diversity enables multi-objective optimization such as low immunogenicity or increased solubility. ProteinMPNN removes a critical bottleneck in design workflows.
Foldit
An online platform where citizens solve protein-folding puzzles as a game. Players’ intuitive manipulations often find conformations that are hard for algorithms to discover. In design mode, users can propose new sequences, some of which have become experimentally active enzymes. Foldit is an early exemplar of citizen science that combines education with research support. The Baker lab continuously supplies new challenges for the community.
vaccine antigen design
A technique that places only the pathogen’s key epitopes on a designed protein scaffold to elicit strong yet safe immune responses. Multicopy display on self-assembling nanocage surfaces enhances B-cell activation. Candidates for SARS-CoV-2 and RSV are already in clinical trials. Computational design selectively emphasizes conserved epitopes, offering broad protection against variants. It is viewed as a rapid-response vaccine platform.