Protein research has grown into a highly interdisciplinary field, touching chemistry, biology, physics, and engineering. Despite major advances in experimental techniques and computational tools, researchers still face a persistent challenge: results from different laboratories are often difficult to compare or reproduce. Addressing this issue, Marc Zimmer, a chemist at Connecticut College, has proposed a coordinated approach that could bring greater consistency to protein science by encouraging the community to adopt a shared set of reference, or “model,” proteins.
Zimmer, M. (2025). Toward an official model protein system, with GFP as an exemplar. Protein Engineering, Design and Selection, 39. https://doi.org/10.1093/protein/gzaf014
Zimmer outlines his proposal in a recent Perspective published in the journal Protein Engineering, Design and Selection. He argues that protein science has reached a level of maturity where informal conventions are no longer sufficient. Instead, the field could benefit from deliberate coordination around a small number of well-characterized proteins that already appear repeatedly in research across different subdisciplines. By formalizing their role, Zimmer suggests, researchers could reduce ambiguity in experimental interpretation and improve the reuse of data.
Marc Zimmer, a chemist at Connecticut College, stated,
“The question Marc Zimmer is addressing is how communities studying similar proteins can benefit by working together.”
The idea takes inspiration from the long-standing use of model organisms in biology. Fruit flies, mice, yeast, and similar systems became powerful tools not only because of their biological relevance, but because research communities collectively invested in shared protocols, databases, and reporting norms. Zimmer contends that protein science lacks an equivalent molecular framework, even though many laboratories independently rely on the same proteins as de facto standards.
Under the proposed model protein system, a limited group of proteins would be paired with shared benchmarks, curated reference datasets, and minimal reporting requirements. The intention is not to limit experimental design, but to ensure that essential parameters are recorded in a way that allows studies to be compared and replicated more easily. Zimmer emphasizes that the selection of model proteins should remain flexible and evolve as research priorities change.
As an initial illustration, he points to proteins that are already deeply embedded in the field, including green fluorescent protein, lysozyme, hemoglobin and myoglobin, RNase A, and bacteriorhodopsin. These molecules have been studied for decades, generating extensive structural, functional, and biophysical data. Their widespread use makes them practical candidates for shared reference points rather than symbolic choices.
Green fluorescent protein provides a particularly clear example of how a model protein can support coordination. Because its fluorescence depends on correct folding, it offers a direct and quantitative indicator of protein function. Its structure remains stable across different organisms, and years of community effort have produced standardized variants, brightness benchmarks, and open datasets. This combination allows results from different laboratories to be compared with relatively little adjustment.
The relevance of shared reference proteins has increased as artificial intelligence and machine learning tools become more common in protein research. Computational models depend heavily on high-quality, comparable data for training and validation. Fluorescent proteins, for instance, are frequently used as benchmark cases because fluorescence offers a straightforward measure of whether a designed protein behaves as intended. Zimmer argues that without agreed-upon standards, the value of such datasets is diminished.
Responses from the research community suggest cautious agreement with the underlying goal. Martin Chalfie, whose work helped establish green fluorescent protein as a foundational research tool, has noted that the core issue is not labeling proteins as models, but encouraging researchers working on similar systems to coordinate their efforts. Others have described the proposal as timely, particularly given growing concerns about reproducibility and data fragmentation.
Zimmer proposes several practical steps to move the idea forward. These include forming a cross-disciplinary steering group, defining transparent criteria for selecting model proteins, creating minimal reporting checklists, and curating reference datasets that can serve as benchmarks for both experimental and computational studies. Such measures, he argues, would help shift effort away from reconciling incompatible methods and toward building on shared results.
For engineering-focused applications, from enzyme optimization to biosensor development, the implications are significant. A common set of reference proteins could streamline validation workflows, improve confidence in performance comparisons, and reduce redundant experimentation. Rather than introducing new tools, Zimmer’s proposal calls for a shared framework that aligns existing practices. If adopted, it could help protein science move toward a more cumulative and reproducible mode of progress.
Adrian graduated with a Masters Degree (1st Class Honours) in Chemical Engineering from Chester University along with Harris. His master’s research aimed to develop a standardadised clean water oxygenation transfer procedure to test bubble diffusers that are currently used in the wastewater industry commercial market. He has also undergone placments in both US and China primarely focused within the R&D department and is an associate member of the Institute of Chemical Engineers (IChemE).
