Biology subjects

Shuai, J.

Publications and source records attributed to Shuai, J..

6 recordsLinked to original sources

bioRxiv2026

Circadian-Modulated Thresholds as a Mechanistic Basis for Sleep-Wake Transitions, Recovery, and Sleepiness

Human sleep-wake cycles arise from the interplay between homeostatic sleep pressure and circadian rhythms, yet the underlying mechanistic basis by which these drives jointly govern state transitions, recovery from sleep loss, and subjective sleepiness remains unclear. Here, we use the extended Phillips-Robinson model incorporating circadian excitation of orexin and locus coeruleus populations, yielding analytically tractable, circadian-modulated thresholds for sleep onset and awakening, and delineating the roles of circadian and homeostatic drives. We show that homeostatic feedback alone generates intrinsic sleep-wake oscillations via a saddle-node on invariant circle bifurcation, while circadian drive reshapes the stability landscape to account for immediate sleep onset and partial first-night recovery after prolonged deprivation, and enables analytic predictions of sleep timing and duration. We further define sleepiness as the distance between the homeostatic state and the active circadian sleep threshold, which robustly predicts subjective sleepiness across various deprivation, restriction, extension, and recovery protocols. Together, these results establish circadian-modulated thresholds as a unifying dynamical principle linking sleep-wake transitions, recovery dynamics, and sleepiness, with implications for circadian misalignment, shift work, and individualized sleep interventions. Author summarySleep and wakefulness arise from the interaction between circadian timing and the gradual accumulation and dissipation of sleep pressure. Although existing computational models have advanced understanding of these processes, most rely on fixed or heuristic rules for switching between sleep and wake states, limiting their ability to explain behavior under sleep deprivation, extension, or restriction. Here, we present a mechanistic description of state switching in which circadian and homeostatic influences act as separable but interacting control dimensions, providing a framework for how circadian modulation shapes state stability, recovery sleep, and subjective sleepiness, without invoking separate mechanisms for normal and perturbed conditions. By recasting sleep regulation in terms of state-dependent dynamical boundaries, our results offer a unified perspective on sleep timing, duration, recovery, and vulnerability to fatigue. This work provides a more physiologically grounded foundation for sleep-wake modeling and a principled basis for understanding circadian misalignment, shift work, jet lag, and other real-world challenges to sleep health.

neuroscience↗

bioRxiv2025

MAIS: an in-vitro sandbox enables adaptive neuromodulation via scalable neural interfaces

Brain-machine interfaces (BMIs) predominantly rely on static digital architectures to decode biological neuronal networks, a paradigm that is incompatible with natural neural coding in the human brain1-4. Bridging this gap is a critical step in combating neuronal dysfunction, enhancing brain functionality, and refining the precision of neuroprosthetics5. The integration of brain organoids with microelectrode array (MEA), as a class of BMIs, offers a humanized in vitro platform with unique biological compatibility advantages for dynamic neuronal decoding. This study resolves the biological-electronic encoding incompatibility of brain organoid-MEA Integration through three progressive breakthroughs. First, a human-machine hybrid agent is developed as a newly proposed bioengineered platform that couples brain organoids together with high-density MEAs and computational chips, enabling closed-loop perturbation of biological neuronal networks via exogenous signals. Second, through plasticity-driven real-time tracking of neuronal activity, we establish dynamically reconfigurable stimulation nodes that self-align with the electrophysiological states of the organoids. This resolves the exogenous-endogenous encoding mismatch by implementing plasticity-driven adaptation principles that ensure biological compatibility through spatially adaptive coordination. Finally, through shared plasticity rules rather than centralized control, we construct the first scalable multi-agent interactive system (MAIS) and demonstrate its real-world applications. Through designed scenarios of pathological/normal neuronal network interaction, we validate that MAIS achieves stable cross-network coordination. MAIS embodies a self-evolving neural coding sandbox in which plasticity-driven dynamic decoding bridges the compatibility gaps between biological and digital systems, providing a scalable and foundational infrastructure for human-centered neural interfaces.

bioengineering↗

bioRxiv2023

Topological design principle for the robustness of necroptosis biphasic, emergent, and coexistent (BEC) dynamics

Biphasic dynamics, the variable-dependent ability to enhance or restrain biological function, is prevalent in natural systems. Accompanied by biphasic dynamics, necroptosis signaling dominated by RIP1 also appears emergent and coexistent dynamics. Here, we identify the RIP1-RIP3-C8 incoherent feedforward loop embedded with positive feedback of RIP3 to RIP1 is the core topology, and the scale-free feature of RIP3 peak value dictates necroptosis BEC dynamics. Entropy production is introduced to quantify the uncertainty of coexistent dynamics. RIP3 auto-phosphorylation is further determined as a complementary process for robustly attaining necroptosis BEC dynamics. Through screening all possible two- and three-node circuit topologies, a complete atlas of three-node circuit BEC dynamics is generated and only three minimal circuits emerge as robust solutions, proving incoherent feedforward loop is the core topology. Overall, through highlighting a finite set of circuits, this study yields guiding principles for mapping, modulating, and designing circuits for BEC dynamics in biological systems.

systems biology↗

bioRxiv2022

SCTC: inference of developmental potential from single-cell transcriptional complexity

Inference of single-cell developmental potential from scRNA-Seq data enables us to reconstruct the pseudo-temporal path of cell development, which is an important and challenging task for single-cell analysis. Single-cell transcriptional diversity (SCTD), measured by the number of expressed genes per cell, has been found to be negatively correlated with the development time, and thus can be considered as a hallmark of developmental potential. However, in some cases, the gene expression level of the cells in the early stages of development may be lower than that of the later stages, which may lead to incorrect estimation of differentiation states by gene diversity-based inference. Here we refer to the economic complexity theory and propose single-cell transcriptional complexity (SCTC) metrics as a measure of single-cell developmental potential, given the intrinsic similarities between biological and economic complex systems. We take into account not only the number of genes expressed by cells, but also the more sophisticated structure information of gene expression by treating the scRNA-seq count matrix as a bipartite network. We show that complexity metrics characterize the developmental potential more accurately than the diversity metrics. Especially, in the early stages of development, cells typically have lower gene expression level than that in the later stages, while their complexity in the early stages is significantly higher than that in the later stages. Based on the measurement of SCTC, we provide an unsupervised method for accurate, robust, and transferable inference of single-cell pseudotime. Our findings suggest that the complexity emerging from the interaction between cells and genes determines the developmental potential, which may bring new insights into the understanding of biological development from the perspective of the complexity theory.

bioinformatics↗

bioRxiv2022

Dear-DIAXMBD: deep autoencoder for data-independent acquisition proteomics

Data-independent acquisition (DIA) technology for protein identification from mass spectrometry and related algorithms is developing rapidly. The spectrum-centric analysis of DIA data without the use of spectra library from data-dependent acquisition (DDA) data represents a promising direction. In this paper, we proposed an untargeted analysis method, Dear-DIAXMBD, for direct analysis of DIA data. Dear-DIAXMBD first integrates the deep variational autoencoder and triplet loss to learn the representations of the extracted fragment ion chromatograms, then uses the k-means clustering algorithm to aggregate fragments with similar representations into the same classes, and finally establishes the inverted index tables to determine the precursors of fragment clusters between precursors and peptides, and between fragments and peptides. We show that Dear-DIAXMBD performs superiorly with the highly complicated DIA data of different species obtained by different instrument platforms. Dear-DIAXMBD is publicly available at https://github.com/jianweishuai/Dear-DIA-XMBD.

bioinformatics↗

bioRxiv2021

DreamDIA-XMBD: deep representation features improve the analysis of data-independent acquisition proteomics

We developed DreamDIA-XMBD, a software suite for data-independent acquisition (DIA) data analysis. DreamDIA-XMBD adopts a data-driven strategy to capture comprehensive information from elution patterns of target peptides in DIA data and achieves considerable improvements on both identification and quantification performance compared with other state-of-the-art methods such as OpenSWATH, Skyline and DIA-NN. More specifically, in contrast to existing methods which use only 6 to 10 selected transitions from spectral library, DreamDIA-XMBD extracts additional features from dozens of theoretical elution profiles originated from different ions of each precursor using a deep representation network. To achieve higher coverage of target peptides without sacrificing specificity, the extracted features are further processed by non-linear discriminative models under the framework of positive-unlabeled learning with decoy peptides as affirmative negative controls. DreamDIA-XMBD is written in Python, and is publicly available at https://github.com/xmuyulab/Dream-DIA-XMBD for high coverage and precision DIA data analysis.

bioinformatics↗