(4m) Targeted Error Correction in Soft and Biological Materials

Motivation:

Biological systems demonstrate robust error correction at all scales, from wound healing in tissues to kinetic proofreading in DNA. Understanding and harnessing these error correcting mechanisms would revolutionize synthetic materials design and provide pathways to novel therapeutics. My research aims to develop principles for targeted error correction in biological systems and synthetic materials, and use these principles to theoretically and computationally design nanomachines that perform targeted error correction.

Proposal:

I will establish a research program centered around targeted error correction that integrates (1) theoretical statistical physics, soft matter, and fluid dynamics with (2) simulation and optimization approaches including inverse design, multiscale modeling, automatic differentiation, and machine learning and (3) builds on foundational ideas in chemical engineering, physics, materials science and biophysics. My research program will be centered around four aims that combine to build a larger vision:

Aim I: We will decode principles of targeted error correction in biology by building mechanistic models of known error correcting biological systems, including DNA mismatch repair, wound closure, and vascular remodeling. We will employ network science, statistical physics, information theory, sticker-spacer models, and colloidal self-assembly to model these systems. Finding common principles across these disparate scales will inform synthetic design of error correcting materials and could lead to new mechanistic understanding of diseases associated with defects in error correction, potentially providing pathways to novel therapeutics.

Aim II: We will discover design principles for selective binding to errors. Many well-understood mechanisms for error correction apply the same mechanism to every component of a system. Instead, we aim to develop error correcting mechanisms that specifically target and bind to abnormalities. We will build up principles of selective binding by designing selectivity in systems of gradually increasing complexity. We will begin by computationally designing a colloidal structure that selectively binds to polytetrahedral colloidal clusters and not to octahedral clusters of equal energy, which will reveal principles about selectivity based on structure alone.

Aim III: We will inverse-design nanomachines with error correcting functions. Biological systems are rife with highly efficient machines that perform complex functions at the nanoscale. To develop similarly complex synthetic error correcting materials for therapeutic and materials applications, we will need to develop principles underlying the design of nanomachines. Building on my expertise in inverse design of complex functions in colloidal systems and Aim I, we will computationally design multistable colloidal structures capable of performing error correcting functions. We will employ differentiable molecular dynamics simulators such as JAX-MD, coupled to enhance sampling algorithms including STRING methods, to perform optimizations over simulations of error correction.

Aim IV: We will map abstract computational models to fine-grained simulations and experiments for closed-loop design. We will find mappings between the abstract designs from Aims I-III, fine-grained simulations (such as patchy particles or coarse-grained force fields such as MARTINI), and experiments. We will make use of machine learning-based approaches in addition to forward simulation. These efforts will benefit from my experience collaborating with experimental groups and my previous work inferring interaction potentials from colloidal trajectory data. As a result, we will propose designs that can directly be translated to experiment, and will develop new methods of extracting information from experimental data.

These four aims will combine to provide insight into biological error correction, propose new therapeutic treatments, and design experimentally-accessible nanomachines that bind to and correct errors in both biological and inorganic systems.

Previous Work:

Simons Junior Fellow, NYU Center for Soft Matter Research & Flatiron Institute (Sept 2023 - present): I am an independent fellow leading a research program on self-organization in biological and inorganic materials. In collaboration with Prof. David Grier at NYU, I discovered a new class of non-equilibrium matter that I term “emergent activity.” In contrast to traditional active matter, individual particles are not driven. Instead, the activity of the particles emerges only as a collective property of their state of organization (King et al, 2024, arXiv:2404.17410). I collaborated with experimentalists at the Center for Soft Matter Research at NYU to experimentally validate my theoretical results. I am additionally collaborating with Prof. Eleni Katifori at Flatiron Institute to create models of the self-organization of vasculature in the placenta (in prep). Abnormal placental chorionic vasculature has been implicated in preterm, diabetic, and preeclamptic pregnancies. By developing the first mechanistic model of chorionic vascular development, I hope to reveal fundamental features of adaptation via vascular remodeling, which my research group will explore further in Aim I.

PhD in Physics, Harvard University; NSF Graduate Research Fellow (Sept 2018 - May 2023): I trained with Michael Brenner, where my research was aimed at the design of bio-inspired functions in self-assembled materials, through computational modeling, optimization and close collaboration with experimental groups. By adapting tools from non-convex optimization and machine learning to physics-based algorithms, I designed qualitatively new materials properties. While traditional self-assembly research has found enormous success in designing static structures, I leveraged automatic differentiation to show that we can design materials to optimize their kinetics: I demonstrated quantitative control over bulk crystallization and transition rates for spherical colloidal particles (Goodrich & King et al, PNAS, 2021). To increase the complexity of the materials we could design, I collaborated with researchers at Google Brain to develop a method for optimizing anisotropic particles. Together, we implemented the method in JAX-MD, an open-source molecular dynamics engine that is fully differentiable. I used this method to design patchy particles that stabilize self-limited assembly (King & Du et al, PNAS, 2024) and undergo designer colloidal reactions (Krueger & King et al, 2024, arXiv:2312.07798). To bridge the gap between these computational results and experimental realizations, I led a collaboration with two experimental groups at Harvard. We developed a particle tracking algorithm for strongly interacting systems (King et al, Physical Review E, 2022) and introduced a method to infer interaction potentials directly from stochastic particle trajectories (King & Engel et al, 2024, arXiv:2406.01522).

Teaching Interests

Philosophy: My experiences in teaching and mentorship have led me to develop a strong vision for guiding students: I will emphasize curiosity, student autonomy, and deep understanding of the material. I am committed to fostering an inclusive learning environment, which I have put into practice both in my teaching experiences as a graduate student and as a mentor for high school students, undergraduates and graduate students.

Experience: For my work as a graduate teaching fellow for physical mathematics, I was awarded a Certificate of Distinction in Teaching. During that course, I helped develop collaborative Python notebooks that demonstrated key ideas from the course, which I found to be well-suited to an active learning environment. My efforts benefitted from a 9-week course I took on teaching pedagogy and practice. I also volunteer as a mentor for the Junior Scientist Internship at BioBus, which recruits students from under-resourced high schools in New York City, and helps them complete entire independent research projects. Additionally, I helped co-found a platform, wowstem.org, that creates videos, blog posts, and lesson plans for middle schoolers about women in STEM throughout history. WOW STEM has received grants from Harvard SEAS and the American Physical Society.

Interests: My experience in teaching, mentoring, and science communication makes me well prepared to teach a variety of courses in chemical engineering, including thermodynamics, fluid mechanics, statistical mechanics, transport phenomena, and mathematical methods. I plan to incorporate computational skill development in my future courses through the use of Google Colab notebooks, which offer free access to a GPU-enabled python environment with minimal setup required.

Selected Honors & Awards

• Simons Foundation Junior Fellowship (Sept 2023 - Present, ~$460,000 over 3 years)
• MIT Rising Stars in Chemical Engineering Workshop (Sept 2024)
• Distinguished Young Scholar Seminar speaker (July 2024)
• Emerging Soft Matter Excellence Award Symposium (Mar 2023)
• Rising Stars in Soft and Biological Materials Symposium (Oct 2022)
• NSF Graduate Research Fellow (2018 - 2023, ~$140,000 over 5 years)
• Certificate of Distinction in Teaching (Fall 2020)

Selected Publications

1. King, E.M.*, Morrell, M.C.*, Sustiel, J.B, Gronert, M., Pastor, H., Grier, D.G. “Scattered waves fuel emergent activity” arXiv 2404.17410.
2. King, E.M.*, Engel, M.C.*, Martin, C.S., Schoenholz, S.S., Manoharan, V.N., Brenner, M.P. “Inferring interaction potentials from stochastic particle trajectories.” arXiv:2406.01522
3. Krueger, R.K.*, King, E.M.*, Brenner, M.P., “Tuning colloidal reactions.” arXiv 2312.07798
4. King, E. M.*, Du, C.X.*, Brenner, M.P. “Programmable patchy particles for materials design.” PNAS. Forthcoming
5. Kimchi, O., King, E. M., and Brenner, M. P. (2023). “Uncovering the mechanism for aggregation in repeat expanded RNA reveals a reentrant transition.” Nature Communications, 14(1), 332.
6. King, E.M., Wang, W., Weitz, D. A., Spaepen, F., and Brenner, M.P. “Correlation Tracking: Using simulations to interpolate highly correlated particle tracks.” Physical Review E 105.4 (2022): 044608.
7. Goodrich, C. P.*, King, E. M.*, Schoenholz, S. S., Cubuk, E. D., and Brenner, M. P. (2021). “Designing self-assembling kinetics with differentiable statistical physics models.” PNAS, 118(10).(2021): e2024083118
8. King, E.M., Gebbie, M. A., and Melosh, N. A. (2019). “Impact of Rigidity on Molecular Self-Assembly.” Langmuir, 35(48), 16062-16069.