Biology subjects

Chen, Z. S.

Publications and source records attributed to Chen, Z. S..

At least 19 recordsLinked to original sources

bioRxiv2026

Toward Optimizing Thalamic Deep Brain Stimulation for Cortical Modulation: A Surrogate Brain Approach

The thalamus is a central hub that interfaces with widespread cortical and subcortical nodes. Thalamic deep brain stimulation (DBS) offers a principled strategy for distributed cortical modulation: since distinct thalamic nuclei project to spatially segregated cortical territories, stimulation at a single thalamic site can influence multiple cortical nodes. Realizing this potential requires accurate subject-specific estimates of directed thalamocortical effective connectivity (EC) and a computational framework for optimizing stimulation parameters that achieve desired cortical responses. Here, we address both challenges using Neural Perturbational Inference (NPI), a surrogate-brain approach that estimates EC by applying virtual perturbations to a nonlinear dynamical model fitted to resting-state fMRI data. We extend NPI to a high-resolution thalamocortical network comprising 360 cortical regions and 442 thalamic voxels spanning 12 nuclei. We introduce two innovations in training: (i) a temporal signal-to-noise ratio (tSNR)-weighted loss accounting for signal heterogeneity, and (ii) a multi-resolution, cross-scale consistency loss that regularizes model complexity. These strategies yield improved performance in synthetic benchmarks across varying tSNR regimes. Leveraging the inferred subject-specific EC, we further formulate a constrained linear control problem to identify sparse thalamic stimulation targets that achieve desired cortical activation patterns. We validate the inferred EC structure on two independent datasets: the MacStim dataset comprising two macaque monkeys with infrared neural stimulation on medial pulvinar, and the HumanTC resting-state fMRI dataset comprising twelve human subjects. Our results reveal site-specific thalamocortical EC profiles, producing interpretable predictions that align with known ground-truth structures. Together, this work establishes a computationally grounded pathway toward personalized optimization of thalamic DBS in both human and nonhuman primates.

neuroscience↗

bioRxiv2026

Hierarchical Gating of Cortical Population Dynamics Drives Pain

The prefrontal cortex and anterior cingulate cortex are key cortical hubs for pain regulation, yet the functional hierarchy between them remains unclear. We examined how the prelimbic cortex (PL) to ACC pathway regulates nociceptive processing and pain behavior in freely moving rats across synaptic, cellular, and network levels. Activation of the PL to ACC pathway reduced aversion to both evoked and spontaneous pain, whereas inhibition increased pain aversion, indicating that this circuit exerts tonic modulatory control. Meanwhile, ex vivo electrophysiology revealed that PL inputs recruit local ACC interneurons to produce feedforward inhibition of ACC pyramidal neurons. At the cellular level, in vivo microendoscopic calcium imaging showed that optogenetic activation of PL axon terminals suppresses nociceptive-evoked activity of ACC pyramidal neurons. At the network level, PL activation reduced pain-induced excitability while centralizing nociceptive information flow within the ACC, resulting in a gated, low output population state. Overall, these findings identify a hierarchically organized cortical circuit that tonically controls pain related sensory and affective experience.

neuroscience↗

bioRxiv2026

A scalable approach to resolving variants of uncertain significance

Over 90% of missense variants across [~]4,000 disease-associated genes are variants of uncertain significance (VUS). Experimental variant effect measurements provide critical evidence about pathogenicity and inform disease biology, but most variants lack data and clinical translation has been limited. The Impact of Genomic Variation on Function Consortium generated experimental data for 62,215 variants across ten genes using multiplexed assays and 1,407 variants across 163 genes using arrayed assays, curated 193,139 additional community-generated variant effect measurements across 30 additional genes, and developed automated calibration methods for translating experimental data and variant effect predictions into clinical evidence. To reduce current VUS, we developed a scalable workflow using only experimental and predictive evidence, enabling reclassification of 75% of the 16,115 VUS in these genes as pathogenic or benign with <1% error. To minimize future VUS, we analyzed >90,000 unobserved variants; 62% had enough evidence to be "preclassified" as pathogenic or benign. We validated our data, evidence and classifications using All of Us and created interactive resources to enable clinical use of the calibrated data. Thus, for 40 genes, representing 1% of the clinical genome, we resolve most existing VUS and future variants, illustrating how systematic use of scalable evidence can empower genomic medicine.

genomics↗

bioRxiv2026

Subspace communication in the hippocampal-retrosplenial axis

The capacity and flexibility of hippocampal circuits for transforming inputs into downstream outputs is fundamental for navigation and memory, yet the circuit-level mechanisms that allow this operation to adapt across experiences remain unknown. We approach this problem by performing large-scale (up to 1024-channel) recordings across the hippocampal-retrosplenial cortex (RSC) circuit in behaving mice, enabling simultaneous access to spiking activity in dentate gyrus (DG), CA3, CA2, CA1, RSC. Based on a linear dimensionality reduction technique known as partial canonical correlation analysis, we identify low-dimensional communication subspaces1 between two regions while accounting for measured third-area influences. These subspaces captured distinct input-output transformations in CA1, linking upstream (DG, CA3, and CA2) hippocampal activity to downstream cortical targets (RSC). Iintrinsic firing properties and anatomical location constrained subspace memberships--members were mapped to deep sublayers of the CA3-CA1-RSC axis during both spatial and non-spatial tasks. These subspaces could recombine overlapping neuronal pools to support distinct interareal interactions across changing experiences and brain states. Reactivation patterns of CA1-CA3 subspaces, but not those of CA1-RSC, during post-experience sleep correlated with replay, reflecting a plasticity-stability balance of the input-output transformation in the hippocampal-retrosplenial axis. Our data suggest a model in which hippocampal-neocortical communication reconfigures predetermined circuit motifs to flexibly encode experiences.

neuroscience↗

bioRxiv2025

Multiplicative couplings facilitate rapid learning and information gating in recurrent neural networks

The mammalian forebrain is the seat of higher cognition with architectural parallels to modern machine learning systems. Specifically, the cortex resembles recurrent neural networks (RNNs) while the thalamus resembles feedforward neural networks (FNNs). How such architectural features endow the forebrain with its learning capacity, is unknown. Here we take inspiration from empirical thalamocortical discovery and develop a multiplicative coupling mechanism between RNN-FNN architectures that collectively enhance their computational strengths and learning. The multiplicative interaction imposes a Hebbian-weight amplification onto synaptic-neuronal coupling, enabling context-dependent gating and rapid switching. We demonstrate that multiplicative feedback-driven synaptic plasticity achieves 2-100 folds of speed improvement in supervised, reinforcement and unsupervised learning settings, boosting memory capacity, model robustness and generalization of RNNs. We further demonstrate the efficacy and biological plausibility of multiplicative gating in modeling multiregional circuits, including a prefrontal cortex-mediodorsal thalamus network for context-dependent decision making, a cortico-thalamic-cortical network for working memory and attention, and an entorhinal cortex-hippocampus network for visuospatial navigation and sequence replay. Taken together, our results demonstrate the profound insights into neuroscience-inspired computation that enable multi-plastic attractor dynamics and computation in recurrent neural circuits.

neuroscience↗

bioRxiv2024

Traveling waves enhance hippocampal-parahippocampal couplings in humanepisodic and working memory

Multiple brain regions in the human medial temporal lobe (MTL) are coordinated in memory processing. Traveling waves is a potential mechanism to coordinate information transfer through organizing the timing or spatiotemporal patterns of wave propagation. Based on direct human intracranial EEG recordings, we detected bidirectional hippocampal and parahippocampal traveling waves (4-10 Hz) along the posterior-anterior axis during a verbal memory task. Hippocampal traveling waves enhanced hippocampal-parahippocampal and intra-hippocampal couplings in both amplitude and phase as well as hippocampal theta phase-gamma amplitude coupling, suggesting a facilitatory role of TWs. Granger causality analysis showed asymmetric information flow, with greater predictability in the parahippocampal-to-hippocampal direction and dominant peak at the beta band (20-30 Hz). Hippocampal power and bidirectional hippocampal-parahippocampal information flow at the gamma band (35-50 Hz) showed reductions during successful memory encoding trials. These results support functional significance of frequency-specific parahippocampal-hippocampal and intra-hippocampal communications during memory encoding and retrieval. HighlightsO_LIBidirectional hippocampal traveling waves enhance hippocampal-parahippocampal couplings in both amplitude and phase. C_LIO_LIHippocampal-parahippocampal gamma coherence is greater in memory retrieval than memory encoding. C_LIO_LIIntra-hippocampal theta phase-gamma amplitude coupling is greater in memory encoding than recall. C_LIO_LIHippocampal power and hippocampal-parahippocampal granger causality in gamma band reduced in successful trials. C_LI

neuroscience↗

bioRxiv2024

Morphological map of under- and over-expression of genes in human cells

Cell Painting images offer valuable insights into a cells state and enable many biological applications, but publicly available arrayed datasets only include hundreds of genes perturbed. The JUMP (Joint Undertaking in Morphological Profiling) Cell Painting Consortium perturbed roughly 75% of the protein-coding genome in human U-2 OS cells, generating a rich resource of single-cell images and extracted features. These profiles capture the phenotypic impacts of perturbing 15,243 human genes, including overexpressing 12,609 genes (using open reading frames, ORFs) and knocking out 7,975 genes (using CRISPR-Cas9). We mitigated technical artifacts by rigorously evaluating data processing options and validated the datasets robustness and biological relevance. Analysis of phenotypic profiles revealed novel gene clusters and functional relationships, including those associated with mitochondrial function, cancer, and neural processes. The JUMP Cell Painting genetic dataset is a valuable resource for exploring gene relationships and uncovering novel functions.

bioinformatics↗

bioRxiv2024

Upregulation of FasII underlies synergistic neuropathological and behavioral defects in a Drosophila model of myotonic dystrophy

Myotonic dystrophy type 1 (DM1) is a multisystemic disorder that has been extensively studied for decades, yet our understanding of its neuropathological aspect remains rudimentary. In this study, we characterized a novel model of DM1 neuropathology by expressing untranslated expanded CUG repeats at the Drosophila larval neuromuscular junction. In this model, both pre- and postsynaptic expression of CUG repeats participate to induce reduction of synaptic boutons, increase of arbor disassembly and impairment of larval locomotor activity. We found that the expression of CUG repeats caused an upregulation of the cell adhesion molecule, FasII (NCAM1 in mammals), in both the motor neurons and the body wall muscles. Knockdown of fasII was sufficient to rescue bouton numbers and locomotor impairment in this model. Further analyses identified the upregulation of the FasII-C isoform as a major contributor of these phenotypes. Remarkably, overexpressing the FasII-A-PEST+ isoform rescued the synaptic and behavioral defects, likely by outcompeting the upregulated FasII-C. Our study provided the foundation for a basic mechanism of synapse dysregulation in DM1.

neuroscience↗

bioRxiv2023

Identifying behavioral links to neural dynamics of multifiber photometry recordings in a mouse social behavior network

Distributed hypothalamic-midbrain neural circuits orchestrate complex behavioral responses during social interactions. How population-averaged neural activity measured by multi-fiber photometry (MFP) for calcium fluorescence signals correlates with social behaviors is a fundamental question. We propose a state-space analysis framework to characterize mouse MFP data based on dynamic latent variable models, which include continuous-state linear dynamical system (LDS) and discrete-state hidden semi-Markov model (HSMM). We validate these models on extensive MFP recordings during aggressive and mating behaviors in male-male and male-female interactions, respectively. Our results show that these models are capable of capturing both temporal behavioral structure and associated neural states. Overall, these analysis approaches provide an unbiased strategy to examine neural dynamics underlying social behaviors and reveals mechanistic insights into the relevant networks.

neuroscience↗

bioRxiv2023

Pixel-wise programmability enables dynamic high-SNR cameras for high-speed microscopy

High-speed wide-field fluorescence microscopy has the potential to capture biological processes with exceptional spatiotemporal resolution. However, conventional cameras suffer from low signal-to-noise ratio at high frame rates, limiting their ability to detect faint fluorescent events. Here, we introduce an image sensor where each pixel has individually programmable sampling speed and phase, so that pixels can be arranged to simultaneously sample at high speed with a high signal-to-noise ratio. In high-speed voltage imaging experiments, our image sensor significantly increases the output signal-to-noise ratio compared to a low-noise scientific CMOS camera ([~]2-3 folds). This signal-to-noise ratio gain enables the detection of weak neuronal action potentials and subthreshold activities missed by the standard scientific CMOS cameras. Our camera with flexible pixel exposure configurations offers versatile sampling strategies to improve signal quality in various experimental conditions.

neuroscience↗

bioRxiv2023

Unbiased single-cell morphology with self-supervised vision transformers

Accurately quantifying cellular morphology at scale could substantially empower existing single-cell approaches. However, measuring cell morphology remains an active field of research, which has inspired multiple computer vision algorithms over the years. Here, we show that DINO, a vision-transformer based, self-supervised algorithm, has a remarkable ability for learning rich representations of cellular morphology without manual annotations or any other type of supervision. We evaluate DINO on a wide variety of tasks across three publicly available imaging datasets of diverse specifications and biological focus. We find that DINO encodes meaningful features of cellular morphology at multiple scales, from subcellular and single-cell resolution, to multi-cellular and aggregated experimental groups. Importantly, DINO successfully uncovers a hierarchy of biological and technical factors of variation in imaging datasets. The results show that DINO can support the study of unknown biological variation, including single-cell heterogeneity and relationships between samples, making it an excellent tool for image-based biological discovery.

bioinformatics↗

bioRxiv2023

Computational models for state-dependent traveling waves in hippocampal formation

Hippocampal theta (4-10 Hz) oscillations have been identified as traveling waves in both rodents and humans. In freely foraging rodents, the theta traveling wave is a planar wave propagating from the dorsal to ventral hippocampus along the septotemporal axis. Motivated from experimental findings, we develop a spiking neural network of excitatory and inhibitory neurons to generate state-dependent hippocampal traveling waves to improve current mechanistic understanding of propagating waves. Model simulations demonstrate the necessary conditions for generating wave propagation and characterize the traveling wave properties with respect to model parameters, running speed and brain state of the animal. Networks with long-range inhibitory connections are more suitable than networks with long-range excitatory connections. We further generalize the spiking neural network to model traveling waves in the medial entorhinal cortex (MEC) and predict that traveling theta waves in the hippocampus and entorhinal cortex are in sink.

neuroscience↗

bioRxiv2023

Neural dynamics in the limbic system during male social behaviors

Sexual and aggressive behaviors are two evolutionarily conserved social behaviors vital for an animals survival and reproductive success. While an increasing number of brain regions in the limbic system have been identified as functionally relevant for these two types of behaviors, an understanding of how social cues are represented across brain regions and how social behaviors are generated via this network activity remains elusive. To gain a holistic view of the neural responses during social behaviors, we utilized multi-fiber photometry to simultaneously record Ca2+ signals of estrogen receptor alpha (Esr1)-expressing cells from 13 limbic brain regions in male mice during sexual and aggressive behaviors and compare the response magnitude and temporal patterns across regions. We find that conspecific sensory information, as well as social action initiation signals, are widely distributed in the limbic system and can be decoded from the network activity. Cross-region correlation analysis reveals striking increases in functional connectivity in the network during the action initiation phase of social behaviors whereas advanced copulation is accompanied by a "dissociated" network state. Based on the response patterns, we propose a mating-biased network (MBN) and an aggression-biased network (ABN) for mediating male sexual and aggressive behaviors, respectively.

neuroscience↗

bioRxiv2022

Computational mechanisms underlying thalamic regulation of prefrontal signal-to-noise ratio in decision making

The mediodorsal (MD) thalamus is a critical partner for the prefrontal cortex (PFC) in cognitive flexibility. Accumulating evidence has shown that the MD regulates task uncertainty in decision making. However, the mechanism of this cognitive process remains unclear. Here we used a reverse-engineering approach and trained biologically-constrained computational models to delineate these mechanisms. We found that the inclusion of an MD-like feedforward module increased robustness to sensory noise, enhanced working memory and enabled rapid context switching in the recurrent PFC network performing two versions of context-dependent decision-making tasks with sensory and mapping uncertainties. Incorporating genetically identified thalamocortical pathways and interneuron cell types replicated neurophysiological findings of neuronal tuning and uncovered attractor-like population dynamics. Our model revealed key computational mechanisms of context-invariant MD in regulating cueing uncertainty and context switching. It also made experimentally testable predictions linking cognitive deficits with disrupted thalamocortical connectivity, prefrontal excitation-inhibition imbalance and dysfunctional inhibitory cell types.

neuroscience↗

bioRxiv2022

Excitatory-Inhibitory Recurrent Dynamics Produce Robust Visual Grids and Stable Attractors

Spatially modulated grid cells has been recently found in the rat secondary visual cortex (V2) during activation navigation. However, the computational mechanism and functional significance of V2 grid cells remain unknown, and a theory-driven conceptual model for experimentally observed visual grids is missing. To address the knowledge gap and make experimentally testable predictions, here we trained a biologically-inspired excitatory-inhibitory recurrent neural network (E/I-RNN) to perform a two-dimensional spatial navigation task with multisensory (e.g., velocity, acceleration, and visual) input. We found grid-like responses in both excitatory and inhibitory RNN units, and these grid responses were robust with respect to the choices of spatial cues, dimensionality of visual input, activation function, and network connectivity. Dimensionality reduction analysis of population responses revealed a low-dimensional torus-like manifold and attractor, showing the stability of grid patterns with respect to new visual input, new trajectory and relative speed. We found that functionally similar receptive fields with strong excitatory-to-excitatory connection appeared within fully connected as well as structurally connected networks, suggesting a link between functional grid clusters and structural network. Additionally, multistable torus-like attractors emerged with increasing sparsity in inter- and intra-subnetwork connectivity. Finally, irregular grid patterns were found in a convolutional neural network (CNN)-RNN architecture while performing a visual sequence recognition task. Together, our results suggest new computational mechanisms of V2 grid cells in both spatial and non-spatial tasks. HighlightsO_LIGrid patterns emerge in trained RNNs with multisensory inputs C_LIO_LIGrid patterns are robust to the RNN input and network connectivity C_LIO_LIPopulation responses show emergent ring-like manifolds and attractors C_LIO_LIGrid-like patterns persist in RNNs while performing a non-spatial task. C_LI

neuroscience↗

bioRxiv2021

Bimodal Remapping of Visual Grids

Spatially modulated neurons from the rat secondary visual cortex (V2) show grid-like firing patterns during freely foraging in open-field enclosures. However, the remapping of the V2 grid cells is not well understood. Here we report two classes of V2 grid cell populations with distinct remapping properties: one regular class with invariant grid field patterns, and the other bimodal class that has remapping induced by environmental manipulations such as changes in enclosure shape, size, orientation and lighting in a familiar environment. The bimodal V2 grid cell pattern remains stable regardless of the follow-up manipulations, but restores to the original firing pattern upon animals re-entry into the familiar environment on the next day or from the novel environment. The bimodal V2 grid cells are modulated with theta frequency during the course of remapping and stabilize quickly. We also found conjunctive bistable V2 grid cells with invariant head directional tuning. Overall, our results suggest a new grid cell mechanism in V2 that is different from the medial entorhinal cortex (MEC) grid cells. HighlightsO_LIBistable V2 grid cells display bimodal or bistable remapping C_LIO_LIV2 grid cell firing patterns are not disrupted in darkness C_LIO_LIV2 grid cells preserve theta frequency modulation during remapping C_LI

neuroscience↗

bioRxiv2021

Egocentric Asymmetric Coding in Sensory Cortical Border Cells

Both egocentric and allocentric representations of space are essential to spatial navigation. Although some studies of egocentric coding have been conducted within and around the hippocampal formation, externally anchored egocentric spatial representations have not yet been fully explored. Here we record and identify two subtypes of border cell in the rat primary somatosensory cortex (S1) and secondary visual cortex (V2). Subpopulations of S1 and V2 border cells exhibit rotation-selective asymmetric firing fields in an either clockwise (CW) or counterclockwise (CCW) manner. CW- and CCW-border cells increase their firing rates when animals move unidirectionally along environmental border(s). We demonstrate that both CW- and CCW-border cells fire in an egocentric reference frame relative to environmental borders, maintain preferred directional tunings in rotated, stretched, dark as well as novel arenas, and switch their directional firings in the presence of multi-layer concentric enclosures. These findings may provide rotation-selective egocentric reference frames within a larger spatial navigation system, and point to a common computational principle of spatial coding shared by multiple sensory cortical areas. HighlightsO_LIEgocentric border cells are present in rat S1 and V2 C_LIO_LISubtypes of border cells display egocentric asymmetric coding C_LIO_LIEgocentric and allocentric streams coexist in sensory cortices C_LIO_LIRotation-selective asymmetric firing is robust with environmental manipulations C_LI

neuroscience↗

bioRxiv2021

A Geometric Framework for Understanding Dynamic Information Integration in Context-Dependent Computation

Prefrontal cortex plays a prominent role in performing flexible cognitive functions and working memory, yet the underlying computational principle remains poorly understood. Here we trained a rate-based recurrent neural network (RNN) to explore how the context rules are encoded, maintained across seconds-long mnemonic delay, and subsequently used in a context-dependent decision-making task. The trained networks emerged key experimentally observed features in the prefrontal cortex (PFC) of rodent and monkey experiments, such as mixed-selectivity, sparse representations, neuronal sequential activity and rotation dynamics. To uncover the high-dimensional neural dynamical system, we further proposed a geometric framework to quantify and visualize population coding and sensory integration in a temporally-defined manner. We employed dynamic epoch-wise principal component analysis (PCA) to define multiple task-specific subspaces and task-related axes, and computed the angles between task-related axes and these subspaces. In low-dimensional neural representations, the trained RNN first encoded the context cues in a cue-specific subspace, and then maintained the cue information with a stable low-activity state persisting during the delay epoch, and further formed line attractors for sensor integration through low-dimensional neural trajectories to guide decision making. We demonstrated via intensive computer simulations that the geometric manifolds encoding the context information were robust to varying degrees of weight perturbation in both space and time. Overall, our analysis framework provides clear geometric interpretations and quantification of information coding, maintenance and integration, yielding new insight into the computational mechanisms of context-dependent computation.

neuroscience↗