Here, I summarize some of my published papers from Ph.D. and postdoc for an academic audience. See the full list at Google Scholar.
Academic Projects
Chromatin enables precise and scalable gene regulation (2022-2024)
Bistability and positive autoregulation in embryonic development (2020-2023)
Cell fate determination is often thought of as a series of binary decisions, where at each branch point a gene is turned ON or OFF. Positive autoregulation, in which a protein activates its own expression, is the simplest network "motif" capable of acting as a binary "switch" whose ON or OFF state is set by the action of transient upstream regulatory factors. Despite the prevalence of positive autoregulatory motifs in nature, very little quantitative work exists to confirm that these motifs actually meet the requirements for bistability, the mathematical property of a dynamical system that allows it to act as a switch. In this work, conducted with Jake Zhao in the lab of Prof. Hernan Garcia at UC Berkeley, we performed live-imaging experiments on living fruit fly embryos to accurately measure all the parameter values in a model of autoregulation for the fushi tarazu (ftz) gene. Our model makes accurate quantitative predictions for ftz expression levels in individual nuclei and shows that the ftz autoregulatory module is indeed bistable. Further analysis in silico suggests that cells decide the binary fate of Ftz in about 30 min.
Dynamics of bistable boundary formation (2019-2021)
In many developmental systems, cells "read" a spatial gradient of some chemical morphogen to determine their position within a tissue. While early models postulated that cells simply detect whether morphogen levels reach a threshold, recent experimental observations show that gene expression boundaries refine over time in a manner inconsistent with these models—and also fail to account for the situation in which morphogen concentrations themselves may change over time. In this work, I put forward two observations about the dynamic behavior of bistable gene expression boundaries. The first observation is that, in a deterministic system with random initial conditions, a gene expression boundary may become more precise over time if the underlying morphogen gradient exhibits a consistent increase or decrease in concentration over both time and space. The second observation is that a bistable chemical network where the involved species diffuse in space can produce a trgene expression boundary that travels with varying speed and direction along a morphogen gradient. Under the proper circumstances this may lead boundaries to stall at precise locations, regardless of where in the tissue the boundary first appears. Both these observations suggest possible means by which developmental systems might achieve reproducible outcomes despite stochastic variation in initial conditions.
Contrasting patterning with optogenetically emulated cell-to-cell signaling (2017-2019)
Intercellular interactions are often difficult to engineer in synthetic biological systems due to the challenges associated with matching chemical parameters. One way to circumvent this difficulty is to simulate communication signals with an externally controlled input such as light. To that end, I collaborated with Dirk Benzinger and Marc Rullan from the lab of Prof. Mustafa Khammash at ETH Zürich to emulate lateral inhibition among S. cerevisiae using single cell-targeting optogenetics. This approach may eventually enable computer-aided control of tissue development and rapid prototyping of intercellular genetic networks.
Prepattern processing by networks of cells (2016-2018)
Throughout embryonic development, gene expression patterns are continuously refined into more complex patterns that guide growth and cell fate in the adult organism. Gaining intuitive insight into how prepatterns are processed into patterns could aid our understanding of extant genetic networks and facilitate the design of synthetic ones. Inspired by the fundamentals of digital signal and image processing, I developed a framework to analyze how networks of interacting cells respond to spatially varying inputs, or prepatterns, such as morphogen gradients or intrinsic parameter variation. My approach incorporates individual cell behavior and network structure to separate the behavior of the cells from the exact shape of the prepattern, providing a high-level way to predict the form of a pattern for arbitrary inputs.
Contrasting patterning in a quorum-sensing system (2016-2018)
Building on the work of Ferreira, Hsia, and Arcak (2014), I collaborated with Mika Tei from the lab of Prof. Adam Arkin to design and implement a spatially distributed lateral inhibition system. In our setup, E. coli interacted via diffusive molecules to repress each other's production of fluorescent protein. We showed how the geometric arrangement of colonies could be used to manipulate the system's ability to generate contrasting fluorescence patterns. Our approach demonstrates that spatially related control parameters may increase design flexibility in synthetic biological applications.
Acoustic interference in bat echolocation (Stanford, 2013-2015)
I performed a theoretical analysis of bat echolocation to investigate how well a bat can pinpoint a target with or without interference from another bat call. I considered the effects of shifting a bat's call in frequency to see if this reduced the effect of interference. My results suggest that significant interference exists between the echolocation calls of bats of the same and of different species, but that frequency shifting may not be the most appropriate method of jamming avoidance for bats with broadband calls.
