Simons Society of Fellows Retreat 2022
The Simons Foundation Society of Fellows returned to the tradition of an annual retreat after a two-year hiatus, this year in the beautiful town of Marana, Arizona, outside of Tucson in the Sonoran Desert. Senior and Junior Fellows gathered to share advances in their research, enjoy the heat, and gather together for communal meals to stimulate scientific discussion and fellowship among members.
The Simons Society of Fellows convenes a unique group of fellows from mathematics, physics and biology in order to facilitate interdisciplinary dialogue. This opportunity challenges the fellows to communicate their work to those outside of their field. Beyond the fellows’ research, the talks also touched on common themes across science, including publication bias, career development and mentorship, fostering conversations that spanned sessions and continued throughout the weekend.
The remote location nurtured organic scientific communication during breaks, when fellows went hiking, mountain biking and swimming against a backdrop of hazy mountains, dazzling sunsets and towering saguaro cactuses. Junior Fellow, Emily Mackevicius, was lucky enough to spot some wild javelinas roaming near the conference center and held the record for spotting 35 distinct species of birds!
On Saturday, the fellows hiked through the Sonoran Desert to see 3,000-year-old petroglyphs left by the ancestors of the O’odham people. Some fellows were lucky enough to try the pleasantly sour fruit of the barrel cactus, and even sight a rarely-seen-above-ground Gila monster.
On our last evening, we shared a special private dinner to celebrate the accomplishments of the fellows. Although we had hoped to stargaze, the clouds had other ideas, so our last night was spent playing games, including a spirited tournament of fast-paced chess.
We are grateful to Emily Klein and Meghan Fazzi for organizing a memorable retreat. Thank you to Marilyn and Jim Simons for supporting and fostering this interdisciplinary scientific community.
Below are brief summaries of 16 talks presented at the annual retreat, 4 from Senior Fellows and 12 from Junior Fellows:
Jarek kicked off the retreat with a discussion of the mathematics underlying machine learning and artificial neural networks. He explained this in the setting of a down-to-earth problem: how could a computer learn to tell the difference between a cat or a dog? He encapsulated the machine learning as three phases. In the first “design” phase, the computer’s human supervisor chooses a mathematical representation of the animals in question. In the second “training” phase, the computer searches for a numerical function of these representations, which tries its best to correctly answer “cat” or “dog” on a given set of training data. Finding the best such function is an optimization problem, and Jarek explained how gradient descent is an effective approach here. Finally, there is an “operations” phase, where the computer takes what it learned in the training phase to try to tell the difference between new cats and dogs.
Emily shared her work on extreme memory specialists. Titmice and black-capped chickadees cache their food, meaning that they hide it to eat at a later time. Caching requires remembering the location of a food item after just a single experience, a skill that impresses those of us with a tendency to lose their keys. Using calcium imaging in freely moving birds, she found that neural activity in the hippocampus reliably encodes and predicts caching behavior, identifying a population of place cells that are distinct from caching cells used to remember the location of a hidden food source. By recording from these neurons, Emily hopes to understand the ensembles that support this extreme memory learning and retrieval.
Raffi is asking the fundamental question, how do cells get their identities? By studying the marvelous moving maggots of D. melanogaster, Raffi is learning how somatosensory neurons develop early in life into cells with distinct morphologies and gene expression profiles that become the complex neural circuit that senses the world. Using transgenic, circuit and molecular neuroscience approaches, Raffi is discovering how a suite of transcription factors, including cut, dictate the developmental pathway a neuron takes, with a special interest in how nitric oxide signaling contributes to sensing stretch.
Jiaoyang told us about exciting new developments in theoretical statistics. We are all familiar with the Gaussian as a “universal distribution” that appears everywhere in nature, thanks to the central limit theorem: the distribution of the means of an independently chosen, large sample always looks like a bell curve. Jiaoyang explained that if we were to instead consider highly correlated samples, we would see a very different family of “universal” distributions. These new distributions arise from the statistics of the eigenvalues of large, random matrices. He told us in particular about the Tracy–Widom distribution, which appears in nature in the growth fronts of forest fires, in liquid crystal formation and in his own work on the statistics of random tiling problems.
Begüm is studying how the development of the nervous system is affected by the environment, using communication between the enteric (gut) nervous system and microbiome as a model. The microbiome is seeded at birth, so she is investigating fundamental questions about when and where gut neurons are born, how they diversify and how they regenerate when microbiota are wiped out. She hopes to understand cellular plasticity by learning about the intimate friendship between our microbiome and our gut neurons.
Kaia presented part of the lesson she prepared for NYC teachers taking the Simons Foundation’s Math for America course about the racist legacy of misuse and misinterpretation of evolutionary biology. She shared a personal story that made her invested in debunking biological claims of race, followed by an introduction to common terms in evolutionary biology and their application to understanding human diversity. At the end of the session, Kaia encouraged thoughtful conversation among the Junior and Senior Fellows about how we can use scientific thought to aid in making our classrooms and labs more anti-racist spaces.
Does that smell funny to you? Joan is using live cell imaging to study gene regulation in olfactory neurons, which have a very unusual organization of the nucleus that allows for these cells to express only one type of receptor. By growing these neurons in a dish, she is able to image active DNA in real time to understand how nuclear organization contributes to gene expression.
Quantum materials are materials whose macroscopic properties arise as emergent phenomena through quantum interactions that are microscopic, or much, much finer. For example, two layers of graphene placed on top of each other at certain “magic” angles become a superconductor. Songtian explained how one of the central challenges in studying these materials is coming up with effective techniques to image them at the extremely small scales at which these quantum interactions happen. Traditional optical microscopy techniques (i.e. shining a laser) are limited by the wavelength of infrared light, and more recent scanning probe techniques (poking the material with extremely fine needles) are painstaking. Songtian’s work combines the two into “scanning near-field optimal microscopy,” or “SNOM,” which promises even more accurate resolutions of these fascinating materials.
A question that has long evaded neuroscientists is how the neurons in our brain manage to survive decades, a major problem given that almost nowhere in the brain makes new neurons in adulthood. To understand their secret to longevity, Moses has been studying how growth factors promote the survival and prevent the death of neurons and gave a historical account as to how this field began. His group’s most recent work challenges decades-old dogma that growth factors keep neurons alive, showing that over time they become less and less dependent on these factors to function. This exciting work opens the opportunity for non-opioid pain management therapies and holds the potential to understand how to keep our brains young and active across the lifespan. At the end of his talk, Moses discussed the importance of preprints and led a discussion about how the publication process could change in the future to help trainees transition to the next stage of their career.
Much of contemporary mathematics is devoted to the study of higher dimensional manifolds — shapes in three dimensions and above, generalizing two-dimensional surfaces like the sphere, the torus and the Möbius band. One of the central challenges is to come up with ways of visualizing and studying these shapes, as humans that fundamentally think in terms of two- or three-dimensional Euclidean space. John shared some different perspectives on some simple, familiar manifolds, like the two-dimensional sphere: there are multiple ways to draw a map of the surface of the Earth, for instance. He explained the fundamental difference between studying the topology — the overall form — of a manifold, and its geometry — measurements on it like length, area, curvature, etc. Despite this, he told us about how they are inextricably linked through theorems that allow us to determine the global topological features of manifolds by integrating local geometric quantities, like the Gauss–Bonnet theorem.
Aleksander’s work is on the question of how to distinguish between abstract higher dimensional manifolds with special structures. He posed a question to us: how could we tell that a torus (the surface of a donut) is different from a sphere? One answer is to think about embedded curves: one can place a rubber band wrapped around the tube and through the hole of the donut that cannot contract without being broken, but every rubber band placed around a sphere will slip right off. A more sophisticated answer is to imagine Maxwell’s equations for an electric field on each surface: the space of solutions has a different dimension in each case. Aleksander told us about how these ideas can be combined to study and attempt to classify certain 6-, 7- and 8-dimensional manifolds by considering some specific field equations arising from physics and studying the singularities of their solutions along certain special embedded submanifolds.
Quantum field theory yields techniques, known broadly under the umbrella of “renormalization,” to understand how the random, quantum behavior of individual particles can give rise to order and structure in large systems. Fedor posed the question of whether this can ever go horribly wrong and leave us with large systems that are chaotic: “can quantum field theory end in a total mess?” Renormalization is governed by a certain ordinary differential equation, which is in itself well behaved; however, Fedor explained that we can introduce chaos through uncertain initial conditions. He gave the example of the Ising model with a complex coupling constant, where the renormalization equation reproduces the well-known Bernoulli map, one of the earliest and most famous examples of a chaotic dynamical system.
A fundamental question in neuroscience is how brains form categories and under what circumstances they generalize. Pooja is studying how monkeys solve tasks, such as telling the number, density or spacing of objects apart. She shows that monkeys generalize across numbers, or in other words, they are more likely to make a mistake with numbers that are closer together. She measures neural activity of the prefrontal cortex during these tasks to understand the main organizing principle of circuit dynamics when monkeys are generalizing and is able to determine the decision a monkey makes by decoding the neural activity in this region. With exciting new behavioral paradigms, Pooja hopes to understand how flexible decision making reorganizes neural activity and the role of the prefrontal cortex in developing abstract knowledge.
A major focus of contemporary number theory has been to find rational number solutions to cubic polynomial equations, or phrased geometrically, finding arithmetically significant points on elliptic curves. A famous result of Mordell–Weil from the mid-20th century states that finitely many such points can be used to generate all the others, and a hugely important question in the subject is to find minimal such generating sets of these special points. Michele explained his own work to define “plectic points,” a new type of special point that can be found using analytical techniques, and his work on proving the conjecture that the plectic points are a minimal generating set for certain famous classes of elliptic curves.
A staggering array of complex mating and courtship behaviors can be found throughout the animal kingdom. Philipp shared his detailed dissection on the courtship behavior of Drosophila fruit flies, where he finds that some species require a very specific context to mate, while others are generalists that will mate under many conditions. As an example, the specialist species Drosophila erecta is observed in the wild to mate in the presence of a single fermenting fruit. By studying these outbred, wild flies in the lab, Philipp discovered that visual and olfactory cues are required to guide courtship. He is now studying the novel dependency on chemosensory cues in D. erecta in order to understand how genetically encoded social behaviors can be shaped by context.
Tony led the group in a discussion of the ethics of non-human animals as research subjects, with a special emphasis on the role of primates in neuroscience research, a hotly debated topic both within and outside neuroscience. Tony explained the importance of studying the brains and behaviors of monkeys to clinical neuroscience: they are much better models of human neural circuits and brain disease than fruit flies, zebra-fish or even other mammals such as rats. We discussed the ethical quandaries involved in studying live monkeys, the importance of measures to ensure their well-being, and how better treatment of the animals can lead to better science. Last, he posed the question of why we don’t extend some of the protections that we afford to our primate relatives to other animals such as mice and rats, which while further away from us evolutionarily may still deserve better protections and treatment.
Carol shared the most recent findings of her laboratory on the molecular and cellular basis of stereoscopic, or binocular, vision. By studying albino mice, which have reduced binocularity compared to pigmented mice, she was able to investigate the genesis and specification of retinal ganglion cells that project to the same side, or opposite side, of the brain, causing binocular vision. The work of her laboratory indicates that albino mice lack stereo vision due to the underdevelopment of retinal ganglion cells in a region of the retina called the ciliary margin zone. This underdevelopment is caused by delayed cell cycling during early embryonic development, leading to abnormal cell shape and connectivity. Remarkably, her lab has identified a drug that can reverse these deficits if given early in development. Her work has broad implications for vision loss in albino humans and provides us insight into how animals integrate information from two eyes, allowing us to see the dynamic three dimensional world that we all navigate daily. Carol ended her talk by discussing mentorship and the role that mentees and mentors have in effective relationships. Her conversation encouraged conversation about how to enhance the training and network of the Junior Fellows in order to help them launch productive, creative careers in their chosen fields.
Friday, March 18
10:00 AM Jarosław Błasiok | Great Ideas in Theoretical Computer Science 10:25 AM Emily Mackevicius | Probing how the hippocampus forms one-shot memories using memory expert birds 10:50 AM Raphael Cohn | The Marvelous Moving Maggot 11:45 AM Jiaoyang Huang | Random Matrix Statistics 12:10 PM Begum Aydin | The role of gut microbiota and immune activity on enteric nervous system development and maintenance 5:25 Kaia Tombak | Race and Evolutionary Biology: The Legacy of Misuse and Misinterpretation 5:25 PM Tony Movshon | Why Monkey Research Matters to Brain Science
Saturday, March 19
4:30 PM Joan Pulupa | Visualizing gene regulation in olfactory sensory neurons using live-cell fluorescence microscopy 4:55 PM Songtian Sonia Zhang | Spatial study of quantum materials with nanolight 5:20 PM Moses Chao | Longevity of neuronal populations 5:55 PM John Morgan | High Dimensional Topology and Geometry
Sunday, March 20
10:00 AM Aleksander Doan | Gauge fields and geometry 10:25 AM Fedor Popov | Can Quantum Field Theory End in a Big Mess? 10:50 AM Pooja Vishwanath | Generalization and abstraction in the brain 11:45 AM Michele Fornea | On remarkable solutions to elliptic curves’ equations 12:10 PM Philipp Brand | Behavioral evolution in a complex world: Of food, friends, and love 5:00 PM Carol Mason | Seeing binocularly: Why albinos lack stereo vision
The Rockefeller University
The role of gut microbiota and immune activity on enteric nervous system development and maintenance
The nervous and immune systems are the main interfaces of the body that can sense and respond to environmental challenges. Interactions between the nervous and immune systems occur under steady-state and inflammatory conditions; however, the mechanisms are not understood. The gastrointestinal (GI) tract is the largest barrier surface where the body is exposed to the environment. This immense surface hosts large amounts of microbes, activated immune cells and the enteric nervous system (ENS), a vast array of neurons with remarkable diversity that control various GI functions. The dynamic interplay between the ENS, gut immune cells and microbiota creates a unique milieu with chronic, yet physiological inflammation. Thus, the ENS of the gut is an ideal system to study neuroimmune interactions upon environmental perturbations due to its proximity to immune cells and chronic exposure to environmental stimuli from the diet, microbiota and pathogens. My research will investigate the role of gut microbiota and immune activity on ENS development and maintenance.
Shortly after birth, a postnatal wave of neurogenesis occurs in the ENS that coincides with microbiota colonization and development of a mature immune system. The mechanisms by which gut microbiota and immune cells affect postnatal enteric nervous system development and maintenance are poorly understood. Our lab described that the perturbation of gut microbes by antibiotics treatment or enteric infection results in the loss of enteric neurons. Interestingly, the reconstitution of “healthy” microbiota induces neuronal recovery in adult mice. However, the mechanisms or signals causing neuronal recovery are not known. Using germ-free and gnotobiotic (mouse models with known bacterial communities) mouse models, I will investigate the effects of microbiota colonization on postnatal ENS development and identify bacterial species that induce neuronal differentiation. In combination with genetic fate-mapping tools, I will identify the origin and mechanisms of neuronal recovery upon post-inflammation microbiota reconstitutions. By doing so, my studies will reveal how microbial perturbations affect neurodevelopment and provide insights into inflammatory autoimmune and neurodegenerative diseases.
The Rockefeller University
Behavioral evolution in a complex world: Of food, friends and love
From dancing birds to singing flies, animals have evolved an astounding diversity of behaviors to attract and choose mating partners in their natural habitats. Guided by innate preferences and aversions, the brain filters complex environments for cues and signals to successfully navigate these reproductive behaviors in appropriate spatial, temporal and social contexts. How does the environment shape sensory evolution and contribute to the diversification of mating behaviors? Using the highly tractable insect nervous system as an inroad, I study the neural mechanisms underlying the evolutionary ecology of reproductive behaviors and how the environment contributes to their diversification.
In Drosophila fruit flies, courtship and mating occurs on fermenting food where many individuals congregate. Replicating this naturalistic context in the lab, I discovered striking differences in the sensory dependence of mating behaviors across species. Most notably, Drosophila erecta requires the presence of food to engage in courtship and only mates in social groups. Using a combination of experimental ethology, in vivo functional imaging, genetic manipulation and optogenetics across species, I revealed that sexual arousal in D. erecta is uniquely gated by food odors that switch the valence of visual stimuli leading to courtship and mating. This provides a possible neural mechanism underlying the environmental modulation of reproductive behaviors, highlighting how social behaviors are shaped by the natural environments in which they evolve.
New York University
Longevity of neuronal populations
Little is known about how neurons become dependent and independent of trophic support. The basis of a switch in dependency for trophic factors is not known. This question is relevant since the inability of mature neurons to maintain or lose resistance to trophic factor deprivation may underlie the development of neurodegenerative diseases.
The Marvelous Moving Maggot
Controlling body movement is among the most essential jobs of animal nervous systems. Maintaining precision and speed of movement despite changes in environment, body size and injury necessitates the use of sensory feedback to adaptively modify motor circuits. Proprioceptive neurons, internal sensors of body position and strain, are particularly well-suited for this task. These neurons must coordinate their growth and positioning with the musculature and epidermis in which they are embedded, presumably through a molecular signal released by these other tissues. However, the body is innervated by many different types of sensory neurons, and their specialized functions likely require distinct responses to such signals.
The body wall of the Drosophila larva contains several types of proprioceptors, as well as neurons that sense touch and pain. This system has proven to be an excellent model for understanding how these different neuron types attain their identities, including distinct dendritic arbors that are specialized for their individual sensory functions. We have identified a signaling pathway that is most strongly active in the proprioceptive neurons and dissected how expression of this pathway is differentially controlled during the development of each type of neuron. Interestingly, this pathway is downstream of Nitric Oxide, a molecule that is known to be released in response to stretch in many tissues. This suggests the intriguing possibility that this pathway may be responsible for coordinating the response of different tissues and neuron types to environmental and developmental perturbations in order to maintain the precise feedback necessary for adaptive motor control.
Gauge fields and geometry
In the 19th century, Maxwell explained electricity, magnetism, and light as effects of the electromagnetic field, whose dynamics is governed by the equations named after him. While Maxwell’s principal goal was to explain natural phenomena, he pointed out that his theory had interesting connections to some purely mathematical problems studied by geometers. For many years, this observation remained merely a curiosity. Today, we understand it as part of a much deeper story, relating geometry to gauge fields: generalizations of the electromagnetic field. The goal of the talk is to explain this story, with a focus on new developments involving Calabi-Yau manifolds, a mysterious class of higher-dimensional geometric shapes which for the last forty years have fascinated mathematicians and physicists alike.
On remarkable solutions to elliptic curves’ equations
I will explain how analytic methods (~calculus) conjecturally produce remarkable algebraic solutions (~fractional numbers) to elliptic curves’ equations: y^2=x^3+Ax+B.
New York University
Random Matrix Statistics
The success of random matrices in modeling physical systems lies in the universality phenomenon of their eigenvalue statistics. The general belief is for systems with a lot of independence, such as a population’s weight distribution, we expect to see the bell curve. However, for systems with many strongly interacting components, we expect to see random matrix statistics. In this talk, I will first discuss some background concerning random matrix statistics. Beyond matrix setting, random matrix statistics are conjectured to govern the asymptotic behavior of various random growth models and interacting particle systems. Next, I will show you some beautiful pictures of various models and discuss recent results proving random matrix statistics for these models.
Probing how the hippocampus forms one-shot memories using memory expert birds
The hippocampus is critical for forming instantaneous (one-shot) memories. What happens in the hippocampus at moments of memory storage and recall? We have developed a strategy to address this question — recording the hippocampus of memory expert birds from the food-caching chickadee family. These birds prolifically hide food items in scattered locations, then return later to retrieve food using memory. We have recorded, using calcium imaging, large populations of neurons from the hippocampus during bouts of food caching and retrieving. I will present evidence for transient memory-related activity in the hippocampus and reactivation of cache-related activity patterns preceding moments of food retrieval. I will compare this transient memory-related activity to sustained changes in hippocampal activity that have been described to occur near locations where an animal consistently receives reward. Finally, I will relate these two modes of hippocampal activity in the context of a predictive model of hippocampus function.
Seeing binocularly: Why albinos lack stereo vision
In higher vertebrates, including humans, retinal ganglion cells (RGCs) extend their axons to the same (ipsilateral) or opposite (contralateral) side of the brain to ensure stereo vision. In albinism, a genetic condition in which pigment (melanin) is disrupted in the skin and eye, the ipsilateral eye-to-brain projection is diminished, altering the architecture of the binocular circuit. Albinism therefore is a good model for investigating the cellular and molecular mechanisms that confer RGC properties during the establishment of the binocular circuit. In the albino retina, compromised competence to generate ipsilateral projecting RGCs is due to the downregulation of cyclin D2, a protein that controls the tempo of the cell cycle. Consequently, albino mice that were also cyclin D2-deficient pigmented exhibit a diminished ipsilateral eye-to-brain projection and compromised depth perception. In albino mice, pharmacological stimulation of calcium channels, known to upregulate cyclin D2 in other cell types, normalizes cyclin D2 expression and the ipsilateral RGC population and improves depth perception. These findings highlight cell cycle regulation by cyclin D2 as critical for the formation and function of the mammalian binocular circuit. They also provide an inroad to understanding how melanin disruption in the retinal pigment epithelium in albinism affects the production of RGCs, a long-standing puzzle.
I will also talk about careers: Tools and guidelines are currently plentiful for navigating your next steps in academia and beyond — how to communicate your science and present yourself, how to apply and interview for faculty or industry positions, on mentoring “up” and “down,” but little on what you do when you get to that next step. Although the Senior Fellows have been informally advising Junior Fellows on these topics in the last years, we will discuss how to formalize this training during the retreat and at our gatherings in New York.
High Dimensional Topology and Geometry
A central theme of mathematics is generalization. Higher dimensional spaces are a generalization of two- and three-dimensional spaces accessible to us by direct visualization. Even though we cannot directly visualize higher dimensional spaces, we can define them precisely, and we can study their (very rich) mathematical structure.
In this talk, I will give some elementary examples of higher dimensional spaces and talk about how mathematicians think about them and gain an understanding of their properties. I will also discuss the relation of and differences between topology (the study of form) and geometry (the study of measurement).
New York University
Why Monkey Research Matters to Brain Science
Neuroscience is based on the experience of physicians treating brain disorders and on the work of scientists studying the brains of animals. The work of physicians helps us to understand the structure of the brain and the mind, but it is work with animals that makes it possible to achieve a mechanistic understanding. Many laymen, and many scientists, view research with animals with distaste or dismay. That goes particularly for research with monkeys, our close evolutionary cousins, which are often the best models for human brain function. Monkey research is under threat in many parts of the world, putting both advances in human knowledge and advances in medical science at risk. Tony Movshon will discuss the benefits of monkey research, the nature and extent of the forces that threaten it, and how thoughtful consideration of the ethical foundations of animal research can help us find our way through.
Probing the neural basis of visually-evoked aggression in Siamese fighting fish
Aggression is a universal behavior that shapes human and animal societies. In humans and other primates, visual information plays a prominent role in eliciting aggression, yet in rodent and fly models commonly used in behavioral neuroscience, aggression is evoked largely by smell and pheromonal cues. Thus, how visual cues influence aggressive behavior — and social behavior more generally — remains relatively unexplored.
Siamese fighting fish (Betta splendens) have been bred for hundreds of generations to select for extreme aggression that is evoked solely by visual cues and is markedly more prominent in males than females. Thus, Betta is an ideal model for studying the neurobiology of visually evoked aggression and how the same socially salient visual cue drives sex differences in behavior. I have developed behavioral assays that reliably elicit aggression in adult Betta, as well as machine learning-based methods to analyze the behavior. I have used molecular methods to identify several brain regions associated with aggression and to determine how these regions intersect with the visual system. Going forward, I am characterizing gene expression in these regions to identify molecular pathways mediating sex differences in behavior and developing physiological methods to examine neural activity in real time.
New York University
Can Quantum Field Theory End in a Big Mess?
We study various quantum field theory with O(N1)×O(N2) symmetry. In studying the behavior of the model in the space of N1 and N2, we find points where zero-Hopf bifurcations occur. In the vicinity of these points, we provide analytical and numerical evidence for the existence of Shilnikov homoclinic orbits. The latter is particular interesting since it induces chaotic behavior in the RG flow. As a simple warm-up example for the study of chaotic RG flows, we also review the non-hermitian Ising chain and show how, for special complex values of the coupling constant, its RG transformations are chaotic and equivalent to the Bernoulli map.
Visualizing gene regulation in olfactory sensory neurons using live-cell fluorescence microscopy
The three-dimensional organization of genetic information in higher organisms is critical for proper gene regulation. Although significant progress has been made in identifying the organizational principles of the genome, little is known about the dynamics of nuclear architecture and how these dynamics interact with gene expression. As biologists, many of the tools that we use provide static information. To study the inherently dynamic process of gene expression, I am employing new live-cell imaging technologies to generate unprecedented temporal resolution.
I have chosen to investigate the most striking example of genome organization leading to gene regulation: olfactory receptor (OR) choice by olfactory sensory neurons (OSNs). In mature OSNs, only one OR gene out of >1000 potential OR genes is expressed, and the process by which the expressed OR is activated is stochastic. OR gene activation is accomplished by the formation of a hub that brings together loci from various chromosomes and associates with and activates a single OR gene. This hub forms under the control of various regulatory proteins including the transcription factors, Lhx2 and Ebf, and the adaptor protein Ldb1. In OSNs, I have fluorescently labeled the genomic loci of the expressed OR and these regulatory proteins, allowing me to monitor their dynamics in living cells. By studying these dynamics, I will gain insight into the interplay between changing genome architecture and gene regulation.
Generalization and abstraction in the brain
Brains are expert general learners, generalizing learning across new stimuli and tasks. This is something deep neural networks struggle to do. In primates, the prefrontal cortex has greatly expanded through evolution. Neural populations in the primate prefrontal cortex exhibit a great range of dynamics to support general learning. We approach our study of what makes brains such experts at generalization by designing behavioral tasks with different demands. My talk will introduce some such tasks and the dynamics we observe in the primate prefrontal cortex neurons during ongoing behavior. I will discuss how these neuronal populations generalize to new information to support decision making and how they can enable abstraction and new learning.
Songtian Sonia Zhang
Spatial study of quantum materials with nanolight
Quantum materials is a rightfully broad term to encompass once disparate fields of physics, centered around ’emergent phenomena.’ While emergent phenomena can be found throughout nature, for example in the oft invoked schools of fish and flights of swallows, the complexity greatly increases when these interactions emerge on atomic and quantum length scales on the order of nanometers. The most direct scientific investigation methods have always been visual — but what happens when the science occurs at a resolution beyond the limits of light itself? This talk will introduce a recent technique known as SNOM (scanning near-field optical microscopy) that allows us to study these emergent phenomena with the requisite nanometer resolution. This has allowed a resurgence of interest in decades-old systems such as superconductors and 2D (two-dimensional) materials, as once inaccessible physics emerge under these new measurement techniques. This talk will explore the role of 2D materials in the continued exploration of even the smallest aspects of quantum materials and the place SNOM occupies in this ever growing scientific endeavor to push our understanding beyond the limits.