Biology subjects

Kim, T. D.

Publications and source records attributed to Kim, T. D..

3 recordsLinked to original sources

bioRxiv2023

Flow-field inference from neural data using deep recurrent networks

Computations involved in processes such as decision-making, working memory, and motor control are thought to emerge from the dynamics governing the collective activity of neurons in large populations. But the estimation of these dynamics remains a significant challenge. Here we introduce Flow-field Inference from Neural Data using deep Recurrent networks (FINDR), an unsupervised deep learning method that can infer low-dimensional nonlinear stochastic dynamics underlying neural population activity. Using population spike train data from frontal brain regions of rats performing an auditory decision-making task, we demonstrate that FINDR outperforms existing methods in capturing the heterogeneous responses of individual neurons. We further show that FINDR can discover interpretable low-dimensional dynamics when it is trained to disentangle task-relevant and irrelevant components of the neural population activity. Importantly, the low-dimensional nature of the learned dynamics allows for explicit visualization of flow fields and attractor structures. We suggest FINDR as a powerful method for revealing the low-dimensional task-relevant dynamics of neural populations and their associated computations.

neuroscience↗

bioRxiv2023

Non-canonical attractor dynamics underlie perceptual decision-making

Perceptual decision-making is the process by which an animal uses sensory stimuli to choose an action or mental proposition. This process is thought to be mediated by neurons organized as attractor networks1,2. However, whether attractor dynamics underlie decision behavior and the complex neuronal responses remains unclear. Here we use simultaneous recordings from hundreds of neurons, together with an unsupervised, deep learning-based method, to discover decision-related neural dynamics in frontal cortex and striatum of rats while the subjects accumulate pulsatile auditory evidence. We found that trajectories evolved along two sequential regimes, the first dominated by sensory inputs, and the second dominated by the autonomous dynamics, with flow in a direction (i.e., "neural mode") largely orthogonal to that in the first regime. We propose that the transition to the second regime corresponds to the moment of decision commitment. We developed a simplified model that approximates the coupled transition in dynamics and neural mode and allows precise inference, from each trials large-scale neural population activity, of a putative neurally-inferred time of commitment ("nTc") on that trial. The simplified model captures diverse and complex single-neuron temporal profiles, such as ramping and stepping3-5, as well as trial-averaged curved trajectories6-8, and reveals distinctions between brain regions. The estimated nTc times were not time-locked to stimulus onset or offset, or to response onset, but were instead broadly distributed across trials. If nTc marks the moment of decision commitment, sensory evidence before nTc should affect the decision, while evidence afterward should not. Behavioral analysis of trials aligned to their estimated nTc times confirmed this prediction. Our results show that the formation of a perceptual choice involves a rapid, coordinated transition in both the dynamical regime and the neural mode of the decision process that corresponds to commitment to a decision, and suggest this moment as a useful entry point for dissecting mechanisms underlying rapid changes in internal state.

neuroscience↗

bioRxiv2023

Functional glycoproteomics by integrated network assembly and partitioning

The post-translational modification (PTM) of proteins by O-linked {beta}-N-acetyl-D-glucosamine (O-GlcNAcylation) is widespread across the proteome during the lifespan of all multicellular organisms. However, nearly all functional studies have focused on individual protein modifications, overlooking the multitude of simultaneous O-GlcNAcylation events that work together to coordinate cellular activities. Here, we describe Networking of Interactors and SubstratEs (NISE), a novel, systems-level approach to rapidly and comprehensively monitor O-GlcNAcylation across the proteome. Our method integrates affinity purification-mass spectrometry (AP-MS) and site-specific chemoproteomic technologies with network generation and unsupervised partitioning to connect potential upstream regulators with downstream targets of O-GlcNAcylation. The resulting network provides a data-rich framework that reveals both conserved activities of O-GlcNAcylation such as epigenetic regulation as well as tissue-specific functions like synaptic morphology. Beyond O-GlcNAc, this holistic and unbiased systems-level approach provides a broadly applicable framework to study PTMs and discover their diverse roles in specific cell types and biological states.

cell biology↗