Simons Society of Fellows Retreat

March 31 – April 3, 2017

The Simons Society of Fellows first annual retreat gathered Junior and Senior Fellows together in Miami, Florida for a diverse program of scientific talks and discussion.

Lectures & Slides

Saturday, April 1

Sunday, April 2

  • James Dama
    Seeking Separability in the Dynamics of Liquids
    (Slides PDF)
  • Tobias Bartsch
    Single-molecule Mechanics of the Molecular Spring that Underlies Hearing
  • Kohei Inayoshi
    Stellar-mass Binary Black Hole Formation as Gravitational Wave Sources
  • Logan Grosenick
    Fast volumetric imaging in mammalian brains
  • Mijo Simunovic
    Building a Human Embryo One Cell at a Time
  • Chervin Laporte
    The Response of the Milky Way to the Large Magellanic Cloud and the Sagittarius Dwarf Spheroidal
  • Michal Breker
    Old tricks-New Playground: Genome-wide Discovery of Essential Pathways in the Plant Superkingdom
  • Tony Movshon
    Representation and computation in sensory and motor systems
    (Slides PDF)


Tim Burbridge

Activity-dependent changes in the early developing brain

Early brain development has an outsized shaping effect on adult circuits and evidence increasingly points to early brain disruptions being linked with eventual developmental disorders. Little is known about the normal patterns of human brain activity before birth and even less is known about the effects of abnormal or disrupted fetal activity. Recording brain activity at even large-scale resolution is extremely difficult and most existing work with animal models has utilized in vitro preparations that may not share the complex patterned spontaneous and early sensory activity of an intact brain. This has led to questions about the existence and relevance of early patterned activity in brain development.

We used calcium (activity) imaging of neurons in awake, early postnatal mice to view large-scale and cellular-level neuronal activity patterns, and combined this with genetic and pharmacological manipulations that let us disrupt specific aspects of patterned brain activity to examine any resulting changes in neuronal circuit architecture. We found that early activity patterns do, in fact, drive circuit-specific activity and refinement in a region-specific manner. In addition, disrupting specific characteristics of activity patterns also had specific and separable effects on individual circuit plasticity, implying that early spontaneous activity patterns are likely tuned to sculpt multiple circuits concurrently.

The ability to record, manipulate, and read out the effects of early patterned activity on circuit development gives us a unique handle into the stereotyped activity patterns that help the brain “wire itself up” during early development. The next step is uncovering which specific neurons and pathways are most affected by these activities and which genetic and molecular programs may be driven through early activity and circuit connectivity patterns. We can now specifically manipulate large and small-scale brain activity in early development as well as the individual cellular responses to that activity, so going forward we will learn about changes in downstream genetic/molecular signaling and/or circuit-plasticity that will better elucidate early activity and its role in brain development and, potentially, in developmental disorder.

Aditi Sheshadri, Columbia University

Dynamical Climate Perturbations and the Fluctuation-dissipation Theorem

Models indicate that much of the dynamical response of the atmosphere to imposed perturbations displays the same spatial patterns as those that dominate temporal fluctuations of the unperturbed climate. Such responses are reminiscent of the predictions of the fluctuation-dissipation theorem, which (amongst other things) relates the perturbed response of a dynamical system to its unperturbed variability. Sheshadri will use a fluctuation-dissipation theorem formulation to show the existence of propagating eigenmodes of the atmosphere, describing systematic poleward migration of wind anomalies, and discuss some of the implications of these propagating modes in climate prediction.

Rafael Maia, Columbia University

From Nano(structure) to Macro(evolution): The Optics of Biological Innovation

Why are some groups of organisms more diverse – both in number of species and in morphology – than others? This fundamental question has guided the study of evolutionary biology over the past centuries, illuminating the relationships between organisms and their environments.

In early stages of diversification, species are often ecologically very similar, differentiating primarily in the ornamental traits used for species recognition and mate attraction. To understand the role these traits have in promoting diversification, Maia’s research focuses on avian plumage, specifically iridescent structural colors. These colors are produced by organized arrays of biomaterials in the feather, such that minute changes to their components can result in entirely novel plumage colors, providing a labile template that can be evolutionarily modified. Maia investigates the intersection between the development of color-producing structures, the physics of their optical function and the evolutionary consequences of morphological innovation in how these colors are produced. By combining these approaches, we can understand how biological constraints can impose limits on diversification, and the impact that key innovations have in shaping the tree of life.

Colin Hill, Columbia University

New Physics from Spectral Distortions of the Cosmic Microwave Background

The energy spectrum of the cosmic microwave background (CMB) is the most perfect blackbody known in nature. This observation remains a crucial underpinning of the hot Big Bang scenario for the birth of the universe. However, the standard model of cosmology predicts small deviations from this blackbody form, which have thus far eluded detection. Hill will discuss the physical origin of these spectral distortions, which are sourced by energy and/or entropy injection into the CMB photon bath at different epochs in cosmological history. He will then present projected spectral distortion constraints from the Primordial Inflation Explorer (PIXIE), a recently proposed NASA satellite mission. These projections include a detailed treatment of the numerous foreground signals that obscure our view of the primordial universe. The PIXIE spectral distortion measurements will yield novel insights into cosmic structure formation, the inflationary paradigm for the origin of the universe, and potentially beyond-the-standard-model particle physics.

Dion Khodagholy, NYU Langone Medical Center

Large-scale Neural Interface Devices

As our understanding of the brain’s physiology and pathology progresses, increasingly sophisticated technologies are required to advance discoveries in neuroscience and develop more effective approaches to treating brain disease. There is a tremendous need for advanced materials solutions at the biotic/abiotic interface to improve the spatiotemporal resolution of neuronal recording. Organic electronic devices offer a unique approach to these challenges, due to their mixed ionic/electronic conduction, mechanical flexibility, enhanced biocompatibility, and capability for drug delivery. We designed, developed, and characterized conformable organic electronic devices in the form of transistors and electrodes to efficiently interface with the brain and acquire neurophysiological activity not previously accessible with recordings from the brain surface. These devices have facilitated large-scale rodent neurophysiology experiments and uncovered a novel hippocampal-cortical oscillatory interaction. The biocompatibility of the devices allowed intra-operative recording from patients undergoing epilepsy surgery, highlighting the translational capacity of this class of neural interface devices. In parallel, we are developing the high-speed electronics and embedded acquisition and storage systems required to make high channel count, chronic neurophysiological recording from animals and human subjects possible.

This multidisciplinary approach will enable the development of new devices based on organic electronics, with broad applicability to the understanding of physiologic and pathologic network activity, control of brain-machine interfaces, and therapeutic closed-loop devices.

Sonja Billerbeck, Columbia University

Synthetic Biology, or How to Teach Baker’s Yeast to Detect Pathogens or Speak a New Language

The goal of synthetic biology is to predictably program single cells or synthetic ecosystems to perform useful, new-to-nature tasks.

The major motivation behind Billerebeck’s research projects during graduate school and postdoctoral work was to explore if and how natural systems can be engineered to perform novel tasks. Her goal was to learn engineering across different levels of biological complexity. To this end, Billerbeck focused on proteins and metabolic pathways during her Ph.D., and continued on to whole cells and synthetic ecosystems during her postdoctoral work.

Billerbeck will present her early postdoctoral work on turning baker’s yeast into an ultra-cheap point-of-care dipstick diagnostic for pathogens as well as on her current work on establishing a new communication language to built synthetic fungal ecosystems. She will close with an outlook into her upcoming work on using synthetic biology approaches to study and (in the future) engineer a complex natural ecosystem – the human mycobiome (all the fungi living in and on us).

Billerbeck hopes by the end of her talk, the audience will appreciate why presumably unrelated efforts like decoding the human genome 25 years ago were a major driving force for her research and get a glimpse of the potentials as well as the risks of engineering and repurposing nature.

Jairo A. Diaz, Center for Soft Matter Research, New York University

DNA-driven Assembly of Liquids at the Mesoscale

Inspired by the remarkable stability exhibited by cell membranes, we stabilize liquid interfaces with high precision using chemically modified DNA strands. By combining surface mobility and packing, we expand conventional free energy landscapes to create unique opportunities for assembling structures capable of surpassing structural arrests.

In his talk, Diaz will present a novel route to guide the self-assembly of liquids at the mesoscale (i.e. ~10 nm- 1 um). He will also share new 3D dimensional structures, such as solid-liquid densely-packed clusters and liquid-liquid crystalline lattices, that can be specifically programmed to emerge under controlled external stimuli (e.g. temperature, pH, etc.).

Herbert Wu, Columbia University

Decision Formation in an Olfactory Delayed Match to Sample Task

Given the same sensory cues, we make very different decisions depending on the context. The ability to make flexible decisions frees us from fixing one action to a specific stimulus, allowing tremendous adaptive advantage in an ever-changing environment. However, the neural mechanism underlying the formation of these decisions is poorly understood.

Wu has developed an odor-guided delayed match to sample task in mice to address this question.

This task represents a context-dependent perceptual decision wherein the first stimulus (sample) is the context in which the animals must interpret the second stimulus (test) to select an action. By recording neural activity in several brain areas, Wu has shown how sensory evidence is gradually transformed into motor action. Further, causal manipulation of a motor area suggests that motor areas do not simply execute action once “decision” is made. Rather, motor areas actively participate in the encoding of contexts and configure how sensory cues are interpreted.

James Stafford, NYU School of Medicine

A Nuclear Perspective of Neurodevelopment: from Pediatric Glioma to Autism Spectrum Disorder

An emerging theme in neurodevelopmental diseases such as autism spectrum disorder is that many mutations and disease mechanisms converge on the nucleus to how a genome is packaged and expressed. These so-called “epigenetic” aberrations often arise at defined time points in development and in specific regions of the nervous system. In fact, certain types of children’s brain tumors fit very similar patterns of developmental origin and epigenetic alterations leading to the hypothesis that some pediatric brain tumors are diseases of neurodevelopment.

In Stafford’s work as a Simons Junior Fellow, he studies select epigenetic alterations to better understand how they give rise to disease and how they might be used to inform clinical outcomes. This work has entailed the development of approaches that can integrate mechanistic biochemistry, proteomics and molecular biology with in vivo disease modeling.

One exciting new avenue of work specifically facilitated by this Junior Fellowship has been the development of an electrode recording approach that allows the combination of precise genetic and molecular manipulation with interrogation of neural network integration and response to environmental stimuli. Together this research is beginning to show that epigenetic lesions arising from distinct neurodevelopmental diseases alter molecular phenotypes, neuronal pathology as well as the functional integration of a neural networks in both common and unique ways.

Andrei Okounkov


Modern ideas in mathematical physics sometimes say something interesting about problems that border on recreational and could have been explained to and appreciated by mathematicians of the XVIII Century. I’ll talk about some results of this kind that have to do with (weighted) counting of partitions, and their higher-dimensional analogs.

Tobias F. Bartsch, Rockefeller University

Single-molecule Mechanics of the Molecular Spring that Underlies Hearing

The astounding sensitivity and dynamic range of mammalian hearing result from the ear’s sensory receptors, the hair cells of the inner ear. Each hair cell features a hair bundle, a cluster of stiff stereocilia whose deflection causes mechanically gated ion channels to open. A long outstanding question in sensory neuroscience concerns the properties and identity of the “gating spring”, the mechanical element that converts bundle deflection to a force capable of opening the channels. A strong candidate for this spring is the tip link, a biopolymer of cadherin molecules that connects pairs of adjacent stereocilia. The link’s compliance results both from the elasticity of individual cadherin domains, which was recently explored by molecular-dynamics simulations, and from the rearrangement of the domains relative to each other, which is stabilized by Ca2+ binding and remains unexplored. Currently no experimental data on the mechanical properties of the tip link exist: hearing involves forces so low that they are difficult to access by traditional experimental methods.

We show that the elasticity of the tip link’s constituents can be measured in the physiologically relevant low-force regime by a single-molecule experiment with a custom-built photonic force microscope. In our assay thermal forces are exploited to sample the proteins’ energy landscapes. Each protein of interest is positioned between a fixed glass substrate and a weakly optically trapped bead. The three-dimensional spatial probability distribution of the bead’s thermal motion is then measured with nanometer precision by a megahertz bandwidth detector. The protein’s energy landscape and elastic properties can be computed directly from the measured spatial probability distribution.

Investigations of the effects of varying the Ca2+ concentration and of pathologically relevant mutations on the proteins’ properties are now in progress.

Joint work with Gilman Dionne, Lawrence Shapiro, Ulrich Müller, and A. J. Hudspeth.

Kohei Inayoshi, Columbia University

Stellar-mass Binary Black Hole Formation as Gravitational Wave Sources

General relativity describes the properties of space-time. One of the most interesting predictions by Einstein’s theory was the existence of gravitational waves (GWs), which are produced by accelerated motion of massive objects. In 2015, Advanced LIGO detected three sources of GWs, which are inferred to be merging binary black holes (BBHs). The origin of such massive and compact BBHs and their formation pathways have been proposed through massive binary evolution and/or stellar dynamics in a dense stellar system.

In his talk, Inayoshi will explain a BBH formation channel via binary evolution of Population III stars, which are the first-generation stars formed at one billion years after the Big Bang. This scenario potentially produce a strong GW background, which is detectable by future LIGO observing run. Next, Inayoshi will propose a new idea of how to distinguish the formation channels of LIGO-BBHs with multi-frequency GW observations.

Logan Grosenick

Fast volumetric imaging in mammalian brains

Recording and understanding the coordinated activity of neurons as they interact in intact mammalian brains is a major challenge in biology that has inspired significant technological innovation. Despite this, fast and sensitive measurement of the high-speed activity of many neurons across an large brain volume in scattering mammalian tissue remains an open problem. I will discuss developments in computational imaging that combine high-dimensional, structured statistics with light field microscopy to allow the synchronous acquisition of single-neuron resolution activity throughout intact tissue volumes as fast as a camera can capture images (currently up to 100 Hz at full camera resolution), as well as approaches to deal with this large volume of streaming data.

Mijo Simunovic, Rockefeller University

Building a Human Embryo One Cell at a Time

The first three weeks of human development are accompanied by dramatic changes in the size and shape of the embryo and the protein expression profile of its cells. The end of the second week of gestation is marked by a process called gastrulation. During gastrulation, which an embryologist Lewis Wolpert anecdotally identified as the most important event in life, cells undergo striking movements that transform a seemingly symmetric cluster of cells into an assembly of distinct and asymmetrically demarcated cell types. This process involves complicated, yet somehow perfectly orchestrated interactions of gene expression regulators that ultimately create three germ layers, each layer later contributing to distinct group of organs. Despite decades of intensive research, we are only beginning to understand the origin of human development, in great part because it is impossible to study human embryos in the same way we study mouse embryos.

We use human embryonic stem cells, micropatterning techniques and gene manipulation to model various aspects of the human embryo. With these techniques, we try to learn what makes us different from other mammals in the timing of early development, when and how cells in the human embryo lose their pluripotency and acquire terminal fates and how the cells organize to form complex and asymmetric structures under external biochemical and mechanical cues. Most recently, we have developed approaches to recreate 3D self-organized structures on a dish that share many important features with the early human embryo. These structures may become a routine synthetic model of early human development.

Chervin Laporte, Columbia University

The Response of the Milky Way to the Large Magellanic Cloud and the Sagittarius Dwarf Spheroidal

The Milky Way disc is asymmetric about the midplane. The HI shows a warp and recent star count surveys show that the disc is rippled (both in height and streaming vertical velocity). Their respective origins are still not settled.
Laporte will show results that suggest that all these features in the disc are shaped by its two most massive satellites. He will present the first self-consistent N-body models of the MW and LMC pair in a first infall scenario designed to study the response of the disc to different mass models. Laporte will show that a massive LMC of virial mass 2.5×10^11 Msun predicts a warp with the correct phases as the observed HI warp, but no ripples.

However, discrepancies remain between the models and the data, and Laporte will discuss some possible reasons. By contrasting the response of the disc with a heavy Sgr model he will also show that no single interaction model (MW+LMC/MW+Sgr) can capture all the features seen in the disc and that one should consider the collective effect of both satellites. Laporte will then present some revised models of Sgr interacting with the MW that he is now using to study the coupling between the LMC and Sgr on the galactic disc (Laporte et al. in prep).

Tony Movshon, NYU

Representation and computation in sensory and motor systems

Only in very special circumstances are the actions of single nerve cells consequential. Brain activity must therefore be understood in terms of the activity of populations of neurons. Recent technological advances have made it possible to record the activity of these populations simultaneously, but we have not yet mastered the challenge of understanding population activity. Dynamical systems approaches can be helpful in revealing the information represented by populations, but it may be more difficult to understand how neural populations compute quantities relevant to behavior.