Biology subjects

Cunningham, J. P.

Publications and source records attributed to Cunningham, J. P..

13 recordsLinked to original sources

bioRxiv2024

Identifying Interpretable Latent Factors with Sparse Component Analysis

In many neural populations, the computationally relevant signals are posited to be a set of latent factors - signals shared across many individual neurons. Understanding the relationship between neural activity and behavior requires the identification of factors that reflect distinct computational roles. Methods for identifying such factors typically require supervision, which can be suboptimal if one is unsure how (or whether) factors can be grouped into distinct, meaningful sets. Here, we introduce Sparse Component Analysis (SCA), an unsupervised method that identifies interpretable latent factors. SCA seeks factors that are sparse in time and occupy orthogonal dimensions. With these simple constraints, SCA facilitates surprisingly clear parcellations of neural activity across a range of behaviors. We applied SCA to motor cortex activity from reaching and cycling monkeys, single-trial imaging data from C. elegans, and activity from a multitask artificial network. SCA consistently identified sets of factors that were useful in describing network computations.

neuroscience↗

bioRxiv2023

Simple decoding of behavior from a complicated neural manifold

Decoders for brain-computer interfaces (BCIs) assume constraints on neural activity, chosen to reflect scientific beliefs while yielding tractable computations. Recent scientific advances suggest that the true constraints on neural activity, especially its geometry, may be quite different from those assumed by most decoders. We designed a decoder, MINT, to embrace statistical constraints that are potentially more appropriate. If those constraints are accurate, MINT should outperform standard methods that explicitly make different assumptions. Additionally, MINT should be competitive with expressive machine learning methods that can implicitly learn constraints from data. MINT performed well across tasks, suggesting its assumptions are well-matched to the data. MINT outperformed other interpretable methods in every comparison we made. MINT outperformed expressive machine learning methods in 37 of 42 comparisons. MINTs computations are simple, scale favorably with increasing neuron counts, and yield interpretable quantities such as data likelihoods. MINTs performance and simplicity suggest it may be a strong candidate for many BCI applications.

bioengineering↗

bioRxiv2021

Flexible neural control of motor units

Voluntary movement requires communication from cortex to the spinal cord, where a dedicated pool of motor units (MUs) activates each muscle. The canonical description of MU function rests upon two foundational tenets. First, cortex cannot control MUs independently but supplies each pool with a common drive. Second, MUs are recruited in a rigid fashion that largely accords with Hennemans size principle. While this paradigm has considerable empirical support, a direct test requires simultaneous observations of many MUs across diverse force profiles. We developed an isometric task that allowed stable MU recordings even during rapidly changing forces. MU activity patterns were surprisingly behavior-dependent. MU activity could not be accurately described as reflecting common drive, but could be captured by assuming multiple drives. Neuropixels probe recordings revealed that, consistent with the requirements of flexible control, the motor cortex population response displays a great many degrees of freedom. Neighboring cortical sites recruited different MUs. Thus, MU activity is flexibly controlled to meet task demands, and cortex may contribute to this ability.

neuroscience↗

bioRxiv2021

Computational Evaluation of DNA Metabarcoding for Universal Diagnostics of Invasive Insect Pests

Appropriate design and selection of PCR primers plays a critical role in determining the sensitivity and specificity of a metabarcoding assay. Despite several studies applying metabarcoding to insect pest surveillance, the diagnostic performance of the short "mini-barcodes" required by high-throughput sequencing platforms has not been established across the broader taxonomic diversity of invasive insects. We address this by computationally evaluating the diagnostic sensitivity and predicted amplification bias for 68 published and novel cytochrome c oxidase subunit 1 (COI) primers on a curated database of 110,676 insect species, including 2,625 registered on global invasive species lists. We find that mini-barcodes between 125-257 bp can provide comparable resolution to the full-length barcode for both invasive insect pests and the broader Insecta, conditional upon the subregion of COI targeted and the genetic similarity threshold used to identify species. Taxa that could not be identified by any barcode lengths were phylogenetically clustered within problem groups, many arising through taxonomic inconsistencies rather than insufficient diagnostic information within the barcode itself. Substantial variation in predicted PCR bias was seen across published primers, with those including 4-5 degenerate nucleotide bases showing almost no mismatch to major insect orders. While not completely universal, a single COI mini-barcode can successfully differentiate the majority of pest and non-pest insects from their congenerics, even at the small amplicon size imposed by 2 x 150 bp sequencing. We provide a ranked summary of high-performing primers and discuss the bioinformatic steps required to curate reliable reference databases for metabarcoding studies.

bioinformatics↗

bioRxiv2021

Partitioning variability in animal behavioral videos using semi-supervised variational autoencoders

Recent neuroscience studies demonstrate that a deeper understanding of brain function requires a deeper understanding of behavior. Detailed behavioral measurements are now often collected using video cameras, resulting in an increased need for computer vision algorithms that extract useful information from video data. Here we introduce a new video analysis tool that combines the output of supervised pose estimation algorithms (e.g. DeepLabCut) with unsupervised dimensionality reduction methods to produce interpretable, low-dimensional representations of behavioral videos that extract more information than pose estimates alone. We demonstrate this tool by extracting interpretable behavioral features from videos of three different head-fixed mouse preparations, as well as a freely moving mouse in an open field arena, and show how these interpretable features can facilitate downstream behavioral and neural analyses. We also show how the behavioral features produced by our model improve the precision and interpretation of these downstream analyses compared to using the outputs of either fully supervised or fully unsupervised methods alone.

neuroscience↗

bioRxiv2021

Learning Sparse Log-Ratios for High-Throughput Sequencing Data

The automatic discovery of sparse biomarkers that are associated with an outcome of interest is a central goal of bioinformatics. In the context of high-throughput sequencing (HTS) data, and compositional data (CoDa) more generally, an important class of biomarkers are the log-ratios between the input variables. However, identifying predictive log-ratio biomarkers from HTS data is a combinatorial optimization problem, which is computationally challenging. Existing methods are slow to run and scale poorly with the dimension of the input, which has limited their application to low- and moderate-dimensional metagenomic datasets. Building on recent advances from the field of deep learning, we present CoDaCoRe, a novel learning algorithm that identifies sparse, interpretable, and predictive log-ratio biomarkers. Our algorithm exploits a continuous relaxation to approximate the underlying combinatorial optimization problem. This relaxation can then be optimized efficiently using the modern ML toolbox, in particular, gradient descent. As a result, CoDaCoRe runs several orders of magnitude faster than competing methods, all while achieving state-of-the-art performance in terms of predictive accuracy and sparsity. We verify the outperformance of CoDaCoRe across a wide range of microbiome, metabolite, and microRNA benchmark datasets, as well as a particularly high-dimensional dataset that is outright computationally intractable for existing sparse log-ratio selection methods.1

bioinformatics↗

bioRxiv2020

Recurrent Switching Dynamical Systems Models for Multiple Interacting Neural Populations

Modern recording techniques can generate large-scale measurements of multiple neural populations over extended time periods. However, it remains a challenge to model non-stationary interactions between high-dimensional populations of neurons. To tackle this challenge, we develop recurrent switching linear dynamical systems models for multiple populations. Here, each high-dimensional neural population is represented by a unique set of latent variables, which evolve dynamically in time. Populations interact with each other through this low-dimensional space. We allow the nature of these interactions to change over time by using a discrete set of dynamical states. Additionally, we parameterize these discrete state transition rules to capture which neural populations are responsible for switching between interaction states. To fit the model, we use variational expectation-maximization with a structured mean-field approximation. After validating the model on simulations, we apply it to two different neural datasets: spiking activity from motor areas in a non-human primate, and calcium imaging from neurons in the nematode C. elegans. In both datasets, the model reveals behaviorally-relevant discrete states with unique inter-population interactions and different populations that predict transitioning between these states.

neuroscience↗

bioRxiv2020

Deep Graph Pose: a semi-supervised deep graphical model for improved animal pose tracking

Noninvasive behavioral tracking of animals is crucial for many scientific investigations. Recent transfer learning approaches for behavioral tracking have considerably advanced the state of the art. Typically these methods treat each video frame and each object to be tracked independently. In this work, we improve on these methods (particularly in the regime of few training labels) by leveraging the rich spatiotemporal structures pervasive in behavioral video -- specifically, the spatial statistics imposed by physical constraints (e.g., paw to elbow distance), and the temporal statistics imposed by smoothness from frame to frame. We propose a probabilistic graphical model built on top of deep neural networks, Deep Graph Pose (DGP), to leverage these useful spatial and temporal constraints, and develop an efficient structured variational approach to perform inference in this model. The resulting semi-supervised model exploits both labeled and unlabeled frames to achieve significantly more accurate and robust tracking while requiring users to label fewer training frames. In turn, these tracking improvements enhance performance on downstream applications, including robust unsupervised segmentation of behavioral "syllables," and estimation of interpretable "disentangled" low-dimensional representations of the full behavioral video. Open source code is available at https://github.com/paninski-lab/deepgraphpose.

animal behavior and cognition↗

bioRxiv2020

Neuroscience Cloud Analysis As a Service

A major goal of computational neuroscience is the development of powerful data analyses that operate on large datasets. These analyses form an essential toolset to derive scientific insights from new experiments. Unfortunately, a major obstacle currently impedes progress: novel data analyses have a hidden dependence upon complex computing infrastructure (e.g. software dependencies, hardware), acting as an unaddressed deterrent to potential analysis users. While existing analyses are increasingly shared as open source software, the infrastructure needed to deploy these analyses - at scale, reproducibly, cheaply, and quickly - remains totally inaccessible to all but a minority of expert users. In this work we develop Neuroscience Cloud Analysis As a Service (NO_SCPLOWEUROC_SCPLOWCAAS): a fully automated analysis platform that makes state-of-the-art data analysis tools accessible to the neuroscience community. Based on modern large-scale computing advances, NO_SCPLOWEUROC_SCPLOWCAAS is an open source platform with a drag-and-drop interface, entirely removing the burden of infrastructure purchase, configuration, deployment, and maintenance from analysis users and developers alike. NO_SCPLOWEUROC_SCPLOWCAAS offers two major scientific benefits to any data analysis. First, NO_SCPLOWEUROC_SCPLOWCAAS provides automatic reproducibility of analyses at no extra effort to the analysis developer or user. Second, NO_SCPLOWEUROC_SCPLOWCAAS cleanly separates tool implementation from usage, allowing for immediate use of arbitrarily complex analyses, at scale. We show how these benefits drive the design of simpler, more powerful data analyses. Furthermore, we show that many popular data analysis tools offered through NO_SCPLOWEUROC_SCPLOWCAAS outperform typical analysis solutions (in terms of speed and cost) while improving ease of use, dispelling the myth that cloud compute is prohibitively expensive and technically inaccessible. By removing barriers to fast, efficient cloud computation, NO_SCPLOWEUROC_SCPLOWCAAS can dramatically accelerate both the dissemination and the effective use of cutting-edge analysis tools for neuroscientific discovery.

neuroscience↗

bioRxiv2020

Value and choice as separable, stable representations in orbitofrontal cortex

Value-based decision-making operates on multiple variables--including offer value, choice, expected outcome, and recent history--each functioning at different times in the decision process. Orbitofrontal cortex (OFC) has long been implicated in value-based decision-making, but it is unclear how downstream circuits might read out complex OFC responses into separate representations of the relevant variables to support different cognitive functions at specific times. We recorded from single neurons in OFC while macaque monkeys made cost-benefit decisions to juice offers. Using a novel analysis--optimal targeted dimensionality reduction--we discovered orthogonal, static dimensions (i.e. linear combinations of neurons) that selectively represented the value, choice, and expected reward of the present and, separately, previous offers. The neural composition of most representations was stable over discrete time periods that aligned to concurrent cognitive demands. We applied a new set of statistical methods to determine that the sensitivity, specificity and stability of the representations were greater than expected from the low-level features--dimensionality and temporal smoothness--of the responses alone. The separability and stability of OFC representations suggest a mechanism by which downstream circuits can read out specific task-relevant variables at appropriate times.

neuroscience↗

bioRxiv2019

Interrogating theoretical models of neural computation with deep inference

1A cornerstone of theoretical neuroscience is the circuit model: a system of equations that captures a hypothesized neural mechanism. Such models are valuable when they give rise to an experimentally observed phenomenon - whether behavioral or a pattern of neural activity - and thus can offer insights into neural computation. The operation of these circuits, like all models, critically depends on the choice of model parameters. A key step is then to identify the model parameters consistent with observed phenomena: to solve the inverse problem. In this work, we present a novel technique, emergent property inference (EPI), that brings the modern probabilistic modeling toolkit to theoretical neuroscience. When theorizing circuit models, theoreticians predominantly focus on reproducing computational properties rather than a particular dataset. Our method uses deep neural networks to learn parameter distributions with these computational properties. This methodology is introduced through a motivational example inferring conductance parameters in a circuit model of the stomatogastric ganglion. Then, with recurrent neural networks of increasing size, we show that EPI allows precise control over the behavior of inferred parameters, and that EPI scales better in parameter dimension than alternative techniques. In the remainder of this work, we present novel theoretical findings gained through the examination of complex parametric structure captured by EPI. In a model of primary visual cortex, we discovered how connectivity with multiple inhibitory subtypes shapes variability in the excitatory population. Finally, in a model of superior colliculus, we identified and characterized two distinct regimes of connectivity that facilitate switching between opposite tasks amidst interleaved trials, characterized each regime via insights afforded by EPI, and found conditions where these circuit models reproduce results from optogenetic silencing experiments. Beyond its scientific contribution, this work illustrates the variety of analyses possible once deep learning is harnessed towards solving theoretical inverse problems.

neuroscience↗

bioRxiv2019

Unsupervised Clusterless Decoding using a Switching Poisson Hidden Markov Model

Spike sorting is a standard preprocessing step to obtain ensembles of single unit data from multiunit, multichannel recordings in neuroscience. However, more recently, some researchers have started doing analyses directly on the unsorted data. Here we present a new computational model that is an extension of the standard (unsupervised) switching Poisson hidden Markov model (where observations are time-binned spike counts from each of N neurons), to a clusterless approximation in which we observe only a d-dimensional mark for each spike. Such an unsupervised yet clusterless approach has the potential to incorporate more information than is typically available from spike-sorted approaches, and to uncover temporal structure in neural data without access to behavioral correlates. We show that our approach can recover model parameters from simulated data, and that it can uncover task-relevant structure from real neural data.

neuroscience↗

bioRxiv2019

Neural trajectories in the supplementary motor area and primary motor cortex exhibit distinct geometries, compatible with different classes of computation

The supplementary motor area (SMA) is believed to contribute to higher-order aspects of motor control. To examine this contribution, we employed a novel cycling task and leveraged an emerging strategy: testing whether population trajectories possess properties necessary for a hypothesized class of computations. We found that, at the single-neuron level, SMA exhibited multiple response features absent in M1. We hypothesized that these diverse features might contribute, at the population level, to avoidance of population trajectory divergence - ensuring that two trajectories never followed the same path before separating. Trajectory divergence was indeed avoided in SMA but not in M1. Network simulations confirmed that low trajectory divergence is necessary when guidance of future action depends upon internally tracking contextual factors. Furthermore, the empirical trajectory geometry - helical in SMA versus elliptical in M1 - was naturally reproduced by networks that did, versus did not, internally track context.

neuroscience↗