Biology subjects

Liu, J. K.

Publications and source records attributed to Liu, J. K..

9 recordsLinked to original sources

bioRxiv2026

Decoding natural scenes from patterned optogenetic responses in mouse visual cortex

A central challenge in developing visual cortical prostheses is to determine how visual stimuli should be transformed into effective patterns of cortical stimulation. Although advances in stimulation technologies, including optogenetics, provide increasingly precise control over cortical activity, it remains unclear whether artificially evoked activity can reproduce the information content of naturally evoked visual representations. Here we establish a quantitative framework for evaluating visual encoding strategies by decoding cortical responses evoked by natural vision and patterned optogenetic stimulation. We developed a novel dual-modal paradigm in awake mice to bridge the gap between endogenous photostimulation and artificial network driving. By co-expressing the high-performance calcium indicator GCaMP6s and the red-shifted, ultra-sensitive opsin rsChRmine-oScarlet in the primary visual cortex (V1), we successfully translated dynamic natural movie frames into patterned, spatiotemporal optogenetic stimulation. Quantitative comparisons of macro-scale dynamics demonstrated that this patterned optogenetic injection evokes cortical states highly comparable and representationally aligned with those driven by actual visual photostimulation. To systematically evaluate the fidelity of these responses, we developed STAR, a deep learning model featuring spatial and temporal attention mechanisms, and successfully reconstructed the frames of natural movies from V1 signals under both experimental modalities. Collectively, our results demonstrate that complex sensory information can be both naturally encoded and synthetically injected into V1 circuits with high decoding fidelity. This work provides an empirical and computational proof-of-concept for intelligent, closed-loop biomimetic encoders, establishing a robust framework for next-generation cortical visual neuroprostheses and bidirectional brain-machine interfaces.

neuroscience↗

bioRxiv2025

Distributed neural dynamics underlie the shift from movement preparation to execution

Dissociating the neural mechanisms of movement preparation from those of execution across the brain is fundamental to understanding motor control. Invasive recordings in primary motor cortex (M1) indicate that shared neural populations support both phases by unfolding within orthogonal subspaces or "manifolds", yet whether and how these dynamics evolve across the broader set of brain regions involved in memory-guided skilled action remains poorly understood. Here, we used magnetoencephalography (MEG) data obtained during a memoryguided delayed finger-sequence task to track fast population dynamics across M1, premotor cortices, and the hippocampus. Linear discriminant analysis decoding of sequence-specific preparatory and execution patterns revealed a transition from preparation to execution patterns across all regions prior to movement onset. Importantly, the onset of execution exhibited region-specific timing, with M1 shifting last,[~] 100 ms before the first button press, consistent with the established cortical motor hierarchy. Low-dimensional trajectories showed that preparatory and execution states occupied distinct manifolds, yet were not fully orthogonal, suggesting partial overlap in tuning. MEG dynamics were dominated by peri-movement phase activity in the primary motor and premotor regions. Nevertheless, above-chance sequence decoding was recoverable from all motor and premotor regions during execution and in M1 during preparation. The state shift and the decoding accuracy were driven by distributed modulations across multiple frequency bands, indicating that full-band population dynamics carry richer information than band-limited features alone. These results demonstrate that non-invasive MEG can resolve continuous, hierarchical neural state transitions in the human brain in the context of memory-guided movement control. They offer mechanistic insight into distributed motor-related brain dynamics and support developing brain-computer-interfaces that incorporate signals from multiple brain regions beyond the primary motor cortex.

neuroscience↗

bioRxiv2025

Scalable spike sorting across thousands of neurons by modeling neural dynamics with NeuroSort

Neural circuit research requires resolving the functional coding mechanisms at the single-neuron scale. However, neural signals captured from extracellular recordings are inherently complex, shaped both by electrode properties and the spatially distributed neural population activity across recording sites. These factors introduce strong signal mixing and waveform variation, making it diffcult for existing spike sorting methods to reliably isolate individual neurons from large-scale recodings. To address this, we develop NeuroSort, which fuses spatiotemporal information through a contrastive learning strategy to adaptively model neural dynamics. NeuroSort is a GPU-enabled, highly efficient spike sorting framework, enabling the isolating of up to thousands of neurons. We validated NeuroSort on in vivo extracellular recordings, including Utah array data from rhesus macaques performing motor tasks and Neuropixels 1.0 probe data from a rat during rest. Moreover, in a recording collected by the 1,024-channel Neuroscroll probe, NeuroSort sorted the largest number of well-isolated neurons from different brain areas while achieving a 2-8 times speedup over other methods. Taken together, NeuroSort generalizes across recording conditions, providing an interpretable and scalable paradigm for whole-brain dynamic analysis.

neuroscience↗

bioRxiv2024

Decoding dynamic visual scenes across the brain hierarchy

Understanding the computational mechanisms that underlie the encoding and decoding of environmental stimuli is a paramount investigation within the domain of neuroscience. Central to this pursuit is the exploration of how the brain represents visual information across its hierarchical architecture. A prominent challenge resides in discerning the neural underpinnings of the processing of dynamic natural visual scenes. Although considerable research efforts have been made to characterize individual components of the visual pathway, a systematic understanding of the distinctive neural coding associated with visual stimuli, as they traverse this hierarchical landscape, remains elusive. In this study, we leverage the comprehensive Allen Visual Coding dataset and utilize the capabilities of deep learning neural network models to study the question of neural coding in response to dynamic natural visual scenes across an expansive array of brain regions. We find that our decoding model adeptly deciphers visual scenes from neural spiking patterns exhibited within each distinct brain area. A compelling observation arises from the comparative analysis of decoding performances, which manifests as a notable encoding proficiency within both the visual cortex and subcortical nuclei, in contrast to a relatively diminished encoding activity within hippocampal neurons. Strikingly, our results reveal a robust correlation between our decoding metrics and well-established anatomical and functional hierarchy indexes. These findings not only corroborate existing knowledge in visual coding using artificial visual stimuli but illuminate the functional role of these deeper brain regions using dynamic natural scenes. Consequently, our results proffer a novel perspective on the utility of decoding neural network models as a metric for quantifying the encoding of dynamic natural visual scenes, thereby advancing our comprehension of visual coding within the complex hierarchy of the brain.

neuroscience↗

bioRxiv2022

Retina Gap Junction Networks Facilitate Blind Denoising in Visual Hierarchy

Gap junctions in the retina are electrical synapses, which strength is regulated byambient light conditions. Such tunable synapses are crucial for the denoising function of the early visual system. However, it is unclear that how the plastic gap junction network processes unknown noise, specifically how this process works synergistically with the brains higher visual centers. Inspired by the electrically coupled photoreceptors, we develop a computational model of the gap junction filter (G-filter). We show that G-filter is an effective blind denoiser that converts different noise distributions into a similar form. Next, since deep convolutional neural networks (DCNNs) functionally reflect some intrinsic features of the visual cortex, we combine G-filter with DCNNs as retina and ventral visual pathways to investigate the relationship between retinal denoising processing and the brains high-level functions. In the image denoising and reconstruction task, G-filter dramatically improve the classic deep denoising convolutional neural network (DnCNN)s ability to process blind noise. Further, we find that the gap junction strength of the G-filter modulates the receptive field of DnCNNs output neurons by the Integrated Gradients method. At last, in the image classification task, G-filter strengthens the defense of state-of-the-arts DCNNs (ResNet50, VGG19 and InceptionV3) against blind noise attacks, far exceeding human performance when noise is large. Our results indicate G-filter significantly enhance DCNNs ability on various blind denoising tasks, implying an essential role for retina gap junction networks in high-level visual processing.

neuroscience↗

bioRxiv2021

Simple Model for Encoding Natural Images by Retinal Ganglion Cells with Nonlinear Spatial Integration

A central goal in sensory neuroscience is to understand the neuronal signal processing involved in the encoding of natural stimuli. A critical step towards this goal is the development of successful computational encoding models. For ganglion cells in the vertebrate retina, the development of satisfactory models for responses to natural visual scenes is an ongoing challenge. Standard models typically apply linear integration of visual stimuli over space, yet many ganglion cells are known to show nonlinear spatial integration, in particular when stimulated with contrast-reversing gratings. We here study the influence of spatial nonlinearities in the encoding of natural images by ganglion cells, using multielectrode-array recordings from isolated salamander and mouse retinas. We assess how responses to natural images depend on first- and second-order statistics of spatial patterns inside the receptive field. This leads us to a simple extension of current standard ganglion cell models. We show that taking not only the weighted average of light intensity inside the receptive field into account but also its variance over space can partly account for nonlinear integration and substantially improve response predictions of responses to novel images. We find that response predictions for salamander ganglion cell classes with large receptive fields profit most from including spatial contrast information. Finally, we demonstrate how this model framework can be used to assess the spatial scale of nonlinear integration. Our results underscore that nonlinear spatial stimulus integration in the salamander retina translates to stimulation with natural images. Furthermore, the introduced model framework provides a simple, yet powerful extension of standard models and may serve as a benchmark for the development of more detailed models of the nonlinear structure of receptive fields. Author SummaryFor understanding how sensory systems operate in the natural environment, an important goal is to develop models that capture neuronal responses to natural stimuli. For retinal ganglion cells, which connect the eye to the brain, current standard models often fail to capture responses to natural visual scenes. This shortcoming is at least partly rooted in the fact that ganglion cells may combine visual signals over space in a nonlinear fashion. We here show that a simple model, which not only considers the average light intensity inside a cells receptive field but also the variance of light intensity over space, can partly account for these nonlinearities and thereby improve current standard models. This provides an easy-to-obtain benchmark for modeling ganglion cell responses to natural images.

neuroscience↗

bioRxiv2020

What does your cell really do? Model-based assessment of mammalian cells metabolic functionalities using omics data

Large-scale omics experiments have become standard in biological studies, leading to a deluge of data. However, researchers still face the challenge of connecting changes in the omics data to changes in cell functions, due to the complex interdependencies between genes, proteins and metabolites. Here we present a novel framework that begins to overcome this problem by allowing users to infer how metabolic functions change, based on omics data. To enable this, we curated and standardized lists of metabolic tasks that mammalian cells can accomplish. We then used genome-scale metabolic networks to define gene modules responsible for each specific metabolic task. We further developed a framework to overlay omics data on these modules to predict pathway usage for each metabolic task. The proposed approach allows one to directly predict how changes in omics experiments change cell or tissue function. We further demonstrated how this new approach can be used to leverage the metabolic functions of biological entities from the single cell to their organization in tissues and organs using multiple transcriptomic datasets (human and mouse). Finally, we created a web-based CellFie module that has been integrated into the list of tools available in GenePattern (www.genepattern.org) to enable adoption of the approach.

systems biology↗

bioRxiv2017

Revealing structure components of the retina by deep learning networks

Deep convolutional neural networks (CNNs) have demonstrated impressive performance on visual object classification tasks. In addition, it is a useful model for predication of neuronal responses recorded in visual system. However, there is still no clear understanding of what CNNs learn in terms of visual neuronal circuits. Visualizing CNNs features to obtain possible connections to neuronscience under-pinnings is not easy due to highly complex circuits from the retina to higher visual cortex. Here we address this issue by focusing on single retinal ganglion cells with a simple model and electrophysiological recordings from salamanders. By training CNNs with white noise images to predicate neural responses, we found that convolutional filters learned in the end are resembling to biological components of the retinal circuit. Features represented by these filters tile the space of conventional receptive field of retinal ganglion cells. These results suggest that CNN could be used to reveal structure components of neuronal circuits.

neuroscience↗

bioRxiv2017

COBRAme: A Computational Framework for Building and Manipulating Models of Metabolism and Gene Expression

Genome-scale models of metabolism and macromolecular expression (ME-models) explicitly compute the optimal proteome composition of a growing cell. ME-models expand upon the well-established genome-scale models of metabolism (M-models), and they enable new and exciting insights that are fundamental to understanding the basis of cellular growth. ME-models have increased predictive capabilities and accuracy due to their inclusion of the biosynthetic costs for the machinery of life, but they come with a significant increase in model size and complexity. This challenge results in models which are both difficult to compute and challenging to understand conceptually. As a result, ME-models exist for only two organisms (Escherichia coli and Thermotoga maritima) and are still used by relatively few researchers. To address these challenges, we have developed a new software framework called COBRAme for building and simulating ME-models. It is coded in Python and built on COBRApy, a popular platform for using M-models. COBRAme streamlines computation and analysis of ME-models. It provides tools to simplify constructing and editing ME-models to enable ME-model reconstructions for new organisms. We used COBRAme to reconstruct a condensed E. coli ME-model called iJL1678b-ME. This reformulated model gives virtually identical solutions to previous E. coli ME-models while using [1/4] the number of free variables and solving in less than 10 minutes, a marked improvement over the 6 hour solve time of previous ME-model formulations. This manuscript outlines the architecture of COBRAme and demonstrates how ME-models can be reconstructed and edited most efficiently using the software.

systems biology↗