Computational Protein Design Then & Now (1988-2024)
The field of protein design, once seen as a “lunatic fringe” problem, has transformed into one of the most revolutionary areas in modern science. During a recent webinar hosted by David Baker from the University of Washington, three distinguished scientists—Bill DeGrado (UCSF), Steve Mayo (Caltech), and Brian Kuhlman (UNC)—reflected on the history, breakthroughs, and future of protein design. Their insights illuminated the incredible progress of this discipline, driven by relentless innovation and collaboration.
Breaking New Ground in Protein Design
At the dawn of protein design, the only known proteins were those shaped by natural evolution. The idea of creating new proteins from scratch seemed implausible. However, three key breakthroughs shifted the paradigm:
- Non-Biological Sequences Can Fold: In 1988, Bill DeGrado and colleagues demonstrated that sequences not found in nature could fold into stable structures. This foundational discovery proved that evolution wasn’t the only path to functional proteins.
- Computational Protein Design: In 1997, Steve Mayo’s lab showed it was possible to design proteins on a computer by determining sequences that could adopt specific structures.
- Flexible Backbone Design: By 2003, Brian Kuhlman and collaborators expanded the scope of protein design to include flexible backbones, enabling the creation of entirely new protein structures.
Bill DeGrado: Simplicity as Sophistication
Bill DeGrado emphasized the importance of simplicity in scientific breakthroughs, quoting Einstein’s dictum, “Make things as simple as possible, but not simpler.” Starting with minimalist peptide designs, his work laid the foundation for creating therapeutic proteins and catalysts. He shared milestones such as the first de novo protein with a native fold and early protein-based therapeutic designs now progressing through clinical trials.
Steve Mayo: Automation and Optimization
Steve Mayo recounted his team’s pioneering efforts to automate protein design, focusing on inverse folding—computing sequences for given backbones. He highlighted the development of thermally stable proteins and their methods for optimizing potential functions. Mayo also addressed the integration of flexibility in design, acknowledging that dynamic protein models remain a challenge.
Brian Kuhlman: Iteration and AI Integration
Brian Kuhlman, who contributed to the design of the first mixed alpha-beta protein, emphasized iterative refinement. By combining sequence design and backbone optimization, his team achieved designs like the highly stable “Top7.” Kuhlman also explored how modern AI methods, such as AlphaFold and diffusion models, are revolutionizing the field, enabling designs with unprecedented precision and scale.
The Power of Community
Throughout the webinar, the speakers acknowledged the collaborative nature of the field. David Baker expressed gratitude to the global protein design community, emphasizing that progress is a shared endeavor. As AI continues to advance, the possibilities for designing novel proteins with real-world applications—such as therapeutics, vaccines, and environmental solutions—are virtually limitless.
Advice for Aspiring Protein Designers
The panelists offered valuable advice for newcomers:
- Join Collaborative Labs: Seek out labs engaged in protein design and take advantage of accessible tools and tutorials.
- Leverage AI: Modern AI tools like AlphaFold and protein MPNN have lowered the barriers to entry.
- Focus on Applications: Combine protein design expertise with impactful fields such as medicine or environmental science.
Looking Ahead
The future of protein design is brimming with promise. While current methods excel at creating stable proteins, the next frontier lies in designing dynamic, functional proteins with diverse applications. With AI enhancing design capabilities and interdisciplinary collaborations accelerating progress, the field is poised to tackle grand challenges, from curing diseases to addressing climate change.
As the webinar concluded, Baker expressed hope that these discussions would inspire a new generation of scientists to build on these foundational discoveries. The story of protein design is far from over, and the next chapter promises to be even more transformative.