Publications

A conserved phase of gametogenesis is the development of oocytes and sperm within cell clusters (germline cysts) that arise through serial divisions of a founder cell. The resulting cell lineage trees (CLTs) exhibit diverse topologies across animals and can give rise to numerous emergent behaviors. Despite their centrality, sex-specific differences underlying the evolution and patterning of these cell trees are unknown. In Drosophila melanogaster, oocytes develop within a highly invariant and maximally branched 16-cell tree whose topology is constrained by the fusome – a branched membranous organelle critical for proper mitosis in females; the same division pattern and topology are widely thought to occur during spermatogenesis. Using highly-resolved three-dimensional reconstructions based on a supervised learning algorithm, we show that cell divisions in male cysts can deviate from the maximally branched pattern, leading to greater topological variability. Furthermore, in contrast to females, fusome fragmentation is common, suggesting germ cell mitoses can occur in its absence. These findings thus add to the repertoire of CLT formation strategies, highlighting the diversity of mechanisms employed during gametogenesis in the animal kingdom.

Quaternions are important for a wide variety of rotation-related problems in computer graphics, machine vision, and robotics. We study the nontrivial geometry of the relationship between quaternions and rotation matrices by exploiting the adjugate matrix of the characteristic equation of a related eigenvalue problem to obtain the manifold of the space of a quaternion eigenvector. We argue that quaternions parameterized by their corresponding rotation matrices cannot be expressed, for example, in machine learning tasks, as single-valued functions: the quaternion solution must instead be treated as a manifold, with different algebraic solutions for each of several single-valued sectors represented by the adjugate matrix. We conclude with novel constructions exploiting the quaternion adjugate variables to revisit several classic pose estimation applications: 2D point-cloud matching, 2D point-cloud-to-projection matching, 3D point-cloud matching, 3D orthographic point-cloud-to-projection matching, and 3D perspective point-cloud-to-projection matching. We find an exact solution to the 3D orthographic least squares pose extraction problem, and apply it successfully also to the perspective pose extraction problem with results that improve on existing methods.

Quaternions are important for a wide variety of rotation-related problems in computer graphics, machine vision, and robotics. We study the nontrivial geometry of the relationship between quaternions and rotation matrices by exploiting the adjugate matrix of the characteristic equation of a related eigenvalue problem to obtain the manifold of the space of a quaternion eigenvector. We argue that quaternions parameterized by their corresponding rotation matrices cannot be expressed, for example, in machine learning tasks, as single-valued functions: the quaternion solution must instead be treated as a manifold, with different algebraic solutions for each of several single-valued sectors represented by the adjugate matrix. We conclude with novel constructions exploiting the quaternion adjugate variables to revisit several classic pose estimation applications: 2D point-cloud matching, 2D point-cloud-to-projection matching, 3D point-cloud matching, 3D orthographic point-cloud-to-projection matching, and 3D perspective point-cloud-to-projection matching. We find an exact solution to the 3D orthographic least squares pose extraction problem, and apply it successfully also to the perspective pose extraction problem with results that improve on existing methods.

We propose Pathfinder, a variational method for approximately sampling from differentiable log densities. Starting from a random initialization, Pathfinder locates normal approximations to the target density along a quasi-Newton optimization path, with local covariance estimated using the inverse Hessian estimates produced by the optimizer. Pathfinder returns draws from the approximation with the lowest estimated Kullback-Leibler (KL) divergence to the true posterior. We evaluate Pathfinder on a wide range of posterior distributions, demonstrating that its approximate draws are better than those from automatic differentiation variational inference (ADVI) and comparable to those produced by short chains of dynamic Hamiltonian Monte Carlo (HMC), as measured by 1-Wasserstein distance. Compared to ADVI and short dynamic HMC runs, Pathfinder requires one to two orders of magnitude fewer log density and gradient evaluations, with greater reductions for more challenging posteriors. Importance resampling over multiple runs of Pathfinder improves the diversity of approximate draws, reducing 1-Wasserstein distance further and providing a measure of robustness to optimization failures on plateaus, saddle points, or in minor modes. The Monte Carlo KL divergence estimates are embarrassingly parallelizable in the core Pathfinder algorithm, as are multiple runs in the resampling version, further increasing Pathfinder's speed advantage with multiple cores.

We employ machine learning to decode the composition of unknown gas mixtures from the output of an array of four electrochemical sensors. The sensors use metal oxide electrodes paired with a ceramic electrolyte, yttria-stabilized zirconia (YSZ), to produce voltage responses to the presence of gases in complex mixtures. The voltages from the sensor array serve as inputs to a machine learning pipeline which first carries out multi-class classification of mixtures into types based on which gases are present at non-zero concentrations, and subsequently predicts gas concentrations given the mixture type. Thus, our model is able to take a single reading from the sensor array in response to gas mixtures involving NO, NO2, C3H8, and NH3, and output a highly accurate prediction of which gases are present in the mixture, along with the concentrations of each constituent gas. Our computational framework can be easily expanded to include additional gases and additional mixture types, allowing it to be used in numerous automotive, industrial and environmental monitoring settings.

Continuum kinetic theories provide an important tool for the analysis and simulation of particle suspensions. When those particles are anisotropic, the addition of a particle orientation vector to the kinetic description yields a dimensional theory which becomes intractable to simulate, especially in three dimensions or near states where the particles are highly aligned. Coarse-grained theories that track only moments of the particle distribution functions provide a more efficient simulation framework, but require closure assumptions. For the particular case where the particles are apolar, the Bingham closure has been found to agree well with the underlying kinetic theory; yet the closure is non-trivial to compute, requiring the solution of an often nearly-singular nonlinear equation at every spatial discretization point at every timestep. In this paper, we present a robust, accurate, and efficient numerical scheme for evaluating the Bingham closure, with a controllable error/efficiency tradeoff. To demonstrate the utility of the method, we carry out high-resolution simulations of a coarse-grained continuum model for a suspension of active particles in parameter regimes inaccessible to kinetic theories. Analysis of these simulations reveals that inaccurately computing the closure can act to effectively limit spatial resolution in the coarse-grained fields. Pushing these simulations to the high spatial resolutions enabled by our method reveals a coupling between vorticity and topological defects in the suspension director field, as well as signatures of energy transfer between scales in this active fluid model.

Neurons from layer II of the entorhinal cortex (ECII) are the first to accumulate tau protein aggregates and degenerate during prodromal Alzheimer’s disease. Here, we use a data-driven functional genomics approach to model ECII neurons in silico and identify the proto-oncogene DEK as a potential driver of tau pathology. By modulating DEK levels in EC neurons in vitro and in vivo, we first validate the accuracy and cell-type specificity of our network predictions. We then show that Dek silencing changes the inducibility of immediate early genes and alters neuron excitability, leading to dysregulation of neuronal plasticity genes. We further find that loss of function of DEK leads to tau accumulation in the soma of ECII neurons, reactivity of surrounding microglia, and eventually microglia-mediated neuron loss. This study validates a pathological gene discovery tool that opens new therapeutic avenues and sheds light on a novel pathway driving tau pathology in vulnerable neurons.

An ensemble of chemical abundances probing different nucleosynthetic channels can be leveraged to build a comprehensive understanding of the chemical and structural evolution of the Galaxy. Using GALAH DR3 data, we seek to trace the enrichment by the supernovae Ia, supernovae II, asymptotic giant branch stars, and neutron-star mergers and/or collapsars nucleosynthetic sources by studying the [Fe/H], [α/Fe], [Ba/Fe], and [Eu/Fe] chemical compositions of ∼50 000 red giant stars, respectively. Employing small [Fe/H]–[α/Fe] cells, which serve as an effective reference-frame of supernovae contributions, we characterize the abundance-age profiles for [Ba/Fe] and [Eu/Fe]. Our results disclose that these age–abundance relations vary across the [Fe/H]–[α/Fe] plane. Within cells, we find negative age–[Ba/Fe] relations and flat age–[Eu/Fe] relations. Across cells, we see the slope of the age–[Ba/Fe] relations evolve smoothly and the [Eu/Fe] relations vary in amplitude. We subsequently model our empirical findings in a theoretical setting using the flexible Chempy Galactic chemical evolution (GCE) code, using the mean [Fe/H], [Mg/Fe], [Ba/Fe], and age values for stellar populations binned in [Fe/H], [Mg/Fe], and age space. We find that within a one-zone framework, an ensemble of GCE model parameters vary to explain the data. Using present day orbits from Gaia EDR3 measurements we infer that the GCE model parameters, which set the observed chemical abundance distributions, vary systematically across mean orbital radii. Under our modelling assumptions, the observed chemical abundances are consistent with a small gradient in the high-mass end of the initial mass function (IMF) across the disc, where the IMF is more top heavy towards the inner disc and more bottom heavy in the outer disc.

This paper describes the first all-sky search for long-duration, quasi-monochromatic gravitational-wave signals emitted by ultralight scalar boson clouds around spinning black holes using data from the third observing run of Advanced LIGO. We analyze the frequency range from 20~Hz to 610~Hz, over a small frequency derivative range around zero, and use multiple frequency resolutions to be robust towards possible signal frequency wanderings. Outliers from this search are followed up using two different methods, one more suitable for nearly monochromatic signals, and the other more robust towards frequency fluctuations. We do not find any evidence for such signals and set upper limits on the signal strain amplitude, the most stringent being ≈10−25 at around 130~Hz. We interpret these upper limits as both an "exclusion region" in the boson mass/black hole mass plane and the maximum detectable distance for a given boson mass, based on an assumption of the age of the black hole/boson cloud system.

The ensemble of chemical element abundance measurements for stars, along with precision distances and orbit properties, provides high-dimensional data to study the evolution of the Milky Way. With this third data release of the Galactic Archaeology with HERMES (GALAH) survey, we publish 678 423 spectra for 588 571 mostly nearby stars (81.2 percent of stars are within <2 kpc), observed with the HERMES spectrograph at the Anglo-Australian Telescope. This release (hereafter GALAH+ DR3) includes all observations from GALAH Phase 1 (bright, main, and faint survey, 70 percent), K2-HERMES (17 percent), TESS-HERMES (5 percent), and a subset of ancillary observations (8 percent) including the bulge and >75 stellar clusters. We derive stellar parameters Teff, log g, [Fe/H], vmic, vbroad, and vrad using our modified version of the spectrum synthesis code Spectroscopy Made Easy (SME) and 1D MARCS model atmospheres. We break spectroscopic degeneracies in our spectrum analysis with astrometry from Gaia DR2 and photometry from 2MASS. We report abundance ratios [X/Fe] for 30 different elements (11 of which are based on non-LTE computations) covering five nucleosynthetic pathways. We describe validations for accuracy and precision, flagging of peculiar stars/measurements and recommendations for using our results. Our catalogue comprises 65 percent dwarfs, 34 percent giants, and 1 percent other/unclassified stars. Based on unflagged chemical composition and age, we find 62 percent young low-α⁠, 9 percent young high-α⁠, 27 percent old high-α⁠, and 2 percent stars with [Fe/H] ≤ −1. Based on kinematics, 4 percent are halo stars. Several Value-Added-Catalogues, including stellar ages and dynamics, updated after Gaia eDR3, accompany this release and allow chrono-chemodynamic analyses, as we showcase.