Intrinsically disordered proteins deviate from the traditional protein paradigm, where function is closely linked to a stable, folded structure. Instead, IDPs exist as dynamic ensembles of conformations, fluctuating constantly rather than settling into a singular architecture. This structural fluidity underpins their critical roles in biological processes such as molecular signaling, cross-linking, and environmental sensing, but it also presents a vexing obstacle for computational modeling. The transient and heterogeneous nature of IDPs means that conventional structure prediction algorithms, which rely on defined folding patterns, falter when applied to these proteins.

To surmount this barrier, the team led by Harvard’s Paulson School of Engineering and Applied Sciences, along with collaborators at Northwestern, leveraged a sophisticated machine learning technique centered around automatic differentiation—a key computational concept widely utilized in deep learning. Automatic differentiation facilitates the calculation of exact derivatives of physical simulations in real-time, enabling precise optimization by highlighting how infinitesimal changes at the amino acid sequence level translate into modifications of ensemble behavior. Essentially, this approach enables a physics-based, gradient-driven search engine for protein sequences, identifying those with specific dynamic properties rather than fixed structures.

This departure from purely data-driven AI models represents a new paradigm: instead of training machine learning systems solely on empirical protein structures, the researchers integrated physics-based molecular dynamics simulations directly into the optimization loop. By doing so, they generated “differentiable” IDPs, whose properties are tethered to the fundamental laws governing molecular interactions and thermal fluctuations. This allows for a rational design of proteins tailored to functions spanning from molecular connectors that form loops to sensors that react to environmental changes.

Ryan Krueger, a graduate student at Harvard and one of the co-lead authors, explained the motivation behind this approach: “We wanted to avoid training models on vast datasets with limited applicability and instead utilize existing, validated simulations to generate new protein designs directly informed by physical reality.” This contrasts with prior strategies that often relied heavily on patterns mined from known protein structures, which are ill-equipped to capture the dynamic heterogeneity of disordered proteins.

The practical implications of this work are profound. IDPs have been implicated in a variety of diseases, notably neurodegenerative disorders such as Parkinson’s disease, where aberrant forms of alpha-synuclein—a prototypical intrinsically disordered protein—contribute to pathology. Being able to design IDPs with targeted functionalities and behaviors opens avenues for not only deeper mechanistic insights but also innovative therapeutic approaches that could modulate or mimic their activity.

From a technical standpoint, the research capitalized on gradient-based optimization methods, routinely used in neural network training, to iteratively refine protein sequences. These methods compute derivatives of objective functions concerning sequence parameters, enabling the algorithm to “climb” toward optimal configurations that exhibit the desired biophysical traits. Unlike heuristic or stochastic search techniques, this ensures computational efficiency and enhanced precision in navigating the vast combinatorial space of amino acid combinations.

Moreover, the team’s strategy integrates seamlessly with molecular dynamics, a computational method that simulates the physical motions of atoms and molecules over time. By coupling automatic differentiation algorithms with these physics-based simulations, the optimization process harnesses the rich dynamic profile of IDPs, including their transient interactions and conformational ensembles, to inform design decisions. This synergistic approach bridges the gap between theoretical modeling and functional protein engineering.

The study, published in the prestigious journal Nature Computational Science, signifies a critical step toward the rational design of biomolecules that elude conventional design frameworks. It comes at a pivotal moment when advances in artificial intelligence are rapidly reshaping biological research, yet intrinsic disorder remains a frontier. The work was co-led by Krishna Shrinivas, an assistant professor at Northwestern University and former NSF-Simons QuantBio Fellow, alongside Michael Brenner, the Catalyst Professor of Applied Mathematics and Applied Physics at Harvard SEAS.

Further supporting this multi-institutional collaboration were federal agencies including the National Science Foundation AI Institute of Dynamic Systems, the Office of Naval Research, and various Harvard-based research centers. The collective expertise spanned applied mathematics, computational physics, and molecular biology, underscoring the interdisciplinary nature essential for tackling such a complex problem.

Looking ahead, the implications of this method extend beyond natural protein systems. In synthetic biology, engineered IDPs designed with specified properties could serve as novel biomaterials, adaptable sensors, or dynamic scaffolds. The ability to computationally tune sequence-ensemble-function relationships with such granularity offers a powerful toolkit for biotechnologists and pharmaceutical developers alike.

In summary, the team’s innovative utilization of automatic differentiation within a physics-based simulation framework provides a robust, data-efficient pathway to unlocking the mysteries of intrinsically disordered proteins. By transcending the limitations of existing AI models, this research sets the stage for designing a previously inaccessible class of proteins with vast biological and clinical potential.

Subject of Research: Not applicable

Article Title: Generalized design of sequence–ensemble–function relationships for intrinsically disordered proteins

News Publication Date: 6-Oct-2025

Web References:

• Article DOI: 10.1038/s43588-025-00881-y
• Associated institutions: Harvard SEAS, Northwestern University

References:

• Original research published in Nature Computational Science

Image Credits: Ramanna Shrinivas

Keywords:
Protein folding, Protein expression, Protein stability, Proteins, Protein activity, Life sciences, Biochemistry, Biomolecules, Machine learning, Artificial neural networks, Deep learning, Computer science, Applied physics, Applied mathematics, Algorithms

At its core, this discovery builds upon years of research into metasurfaces, which emerged as a pivotal solution to the limitations imposed by traditional optical systems that rely on bulky lenses. These lightweight devices, characterized by their nanoscale structures, have proven invaluable in a plethora of applications, allowing for precise control over light behavior at wavelengths previously deemed unattainable. The transition from conventional optics to metasurfaces marks a significant turning point in optical engineering and design, offering potential solutions for advancements in communications and imaging technologies.

The recent innovation—the bilayer metasurface—is notable for featuring two distinct layers of titanium dioxide nanostructures, effectively doubling the design’s capabilities. Under microscopic examination, the new structure resembles a cityscape of miniature skyscrapers, reflecting the intricate engineering that underpins its functionality. More than just an aesthetic improvement, this layering enables enhanced control over light’s properties such as wavelength, phase, and polarization, pushing the boundaries of what is possible with light manipulation.

Federico Capasso, the senior author of the study, articulates the importance of this development, stating that it represents a pinnacle achievement in nanotechnology. The bilayer design signifies a shift towards more sophisticated optical solutions, allowing researchers to explore dual functionalities within a single device. For example, these meta-optical systems could theoretically project one vivid image from one side and entirely different information from the opposite, showcasing a transformative approach to how visual data is conveyed through optical means.

The evolution from single-layer to bilayer metasurfaces addresses inherent constraints found in earlier models, particularly in how they handle light polarization. Historically, single-layer metasurfaces necessitated specific conditions to manipulate light’s polarization effectively. The flexibility introduced by the bilayer design paves the way for more complex optical devices without the need for intricate setups, making them vastly more practical for a range of applications.

Creating the bilayer metasurface demanded an unprecedented level of precision and innovation in fabrication techniques. Researchers utilized the facilities at Harvard’s Center for Nanoscale Systems to pioneer a fabrication process for robust, freestanding structures that maintain chemical independence between the two layers. This multi-level fabrication approach is a significant technical achievement, bridging the gap between advanced nanostructures and their practical optical applications.

The significance of this advancement is further underscored by its potential applications. With their newfound capabilities, these bilayer metasurfaces may facilitate the development of devices that integrate various functionalities into one compact unit, thus streamlining several aspects of optical engineering. The potential to create multifunctional optical devices signifies a monumental step towards sophisticated design in the field of optics, promoting innovation in commercial applications such as smartphone cameras, virtual reality headsets, and more.

Capasso and his team validated their groundbreaking design by demonstrating its functionality in a controlled experiment that manipulated polarized light. This proof-of-concept not only showcases the efficacy of their bilayer system but also hints at a future where additional layers may further enhance control over light. The prospects of incorporating more layers into metasurface designs open the door to extreme broadband operations with both high efficiency and precision, enhancing the potential for even more intricate optical functionalities.

The research team employed various federal funding sources to support this ambitious endeavor, with backing from the Office of Naval Research and the Air Force Office of Scientific Research, among others. This collaborative support highlights the broader implications of their work, emphasizing its significance in both academic and applied contexts. Moreover, the deployment of their technology through partnerships, such as with Harvard’s Office of Technology Development and the establishment of Metalenz, signifies a trajectory aimed at expert spin-off ventures that aspire to commercialize breakthrough optical technologies.

In addition to the compelling scientific advancements, this work exemplifies how interdisciplinary collaboration drives progress in complex fields like nanotechnology. In their pursuit of innovation, the team relied on contributions from various experts to refine their fabrication methods, reiterating the importance of collaboration in today’s research landscape. Such teamwork fosters growth, ensuring that cutting-edge technologies can transition swiftly from research labs to real-world applications.

In conclusion, the introduction of the bilayer metasurface represents a pivotal moment in optical science, and it may herald a new era of technology that enhances our capability to understand and utilize light. From augmented reality to advanced imaging systems, the implications of this research are vast and varied. As further studies unfold and newer applications are envisioned, the ability to manipulate light at unprecedented levels promises to facilitate changes across myriad scientific fields. The advancements driven by this technology will surely stimulate curiosity and further investigation into the exciting world of optical engineering.

Subject of Research:
Article Title: Free-standing bilayer metasurfaces in the visible
News Publication Date: 1-Apr-2025
Web References: https://www.nature.com/articles/s41467-025-58205-7
References: https://seas.harvard.edu/news/2016/06/metalens-works-visible-spectrum-sees-smaller-wavelength-light
Image Credits: Credit: Capasso group / Harvard SEAS

Keywords

Metasurfaces, visible light, light sources, nanostructures, light polarization, fabrication, applied physics, engineering, materials engineering, materials processing, microstructures, optics, nonlinear optics, optical properties, quantum optics.