Biology subjects

Wang, L.

Publications and source records attributed to Wang, L..

At least 145 records · Page 8Linked to original sources

bioRxiv2026

Efficient and accurate near telomere-to-telomere haplotype reconstruction of diploid genomes

Telomere-to-telomere (T2T) and haplotype-resolved assembly are crucial for understanding eukaryotic genomes. For diploid species, this resolution is critical to uncover allelic variations, inheritance patterns, and functional genomic traits. Current scaffolding methods typically employ either sequence-based or graph-based strategies. Sequence-based approaches rely on proximity signals to yield high contiguity, but underutilize assembly graph information, resulting in more structural errors and chromosomal misassignments. Graph-based methods leverage graph topology for higher accuracy but frequently struggle to achieve chromosome-scale contiguity. However, neither strategy alone can overcome its inherent limitations to simultaneously achieve high contiguity and accuracy. To address these challenges, we introduce HapFold, the first hybrid scaffolding framework that synergistically leverages the complementary strengths of both graph-based and sequence-based approaches. By integrating the topological accuracy of assembly graphs with the proximity-guided contiguity of sequence models, HapFold achieves highly accurate, chromosome-scale or near-T2T haplotype reconstructions for diploid genomes. Compared to existing methods, HapFold achieves superior assembly quality while accelerating computation by an order of magnitude. Furthermore, in the haplotype reconstruction of diploid genomes using standard Oxford Nanopore Technologies simplex reads, HapFold enables the reconstruction of a greater number of near-T2T assemblies. Our approach provides a robust and scalable solution for the high-fidelity reconstruction of haplotype-resolved diploid genomes.

bioinformatics↗

bioRxiv2026

Genos-m: a foundation model for human-associated microbial genomes

Human-associated microbial genomes encode extensive strain-level diversity and niche-specific gene repertoires that are critical to host health. However, these complex sequence features remain difficult to capture using general-purpose DNA foundation models, highlighting the need for dedicated representation learning tailored to the human microbiome. Here, we introduce Genos-m, an open-source foundation model for human-associated microbial genome representation. Genos-m was pretrained on approximately 1.2 trillion nucleotide tokens from a curated microbial genome corpus, including human-associated prokaryotic isolates, high-quality metagenome-assembled genomes (MAGs) and bacteriophages, supplemented with GTDB species-level representative genomes to broaden prokaryotic taxonomic breadth. The model uses a sparsely activated Mixture-of-Experts (MoE) Transformer architecture, with 4.7 billion total parameters, approximately 330 million activated parameters per forward pass and a maximum context length of one million base pairs. We evaluated frozen Genos-m representations across short-sequence and gene-level tasks, biosynthetic gene cluster (BGC)-based regional sequence tasks, whole-genome strain phenotype prediction, and zero-shot transfer on prokaryote-related RNAfitness assays. Across these benchmarks, Genos-m consistently ranked among the leading comparison models, with the best performance in five of eight gene-fitness regression tasks and in BGC type classification. Using sparse autoencoders, we identified sparse features in Genos-m hidden activations that aligned with annotated ORFs, intergenic regions, and tRNA and rRNA loci. In downstream applications, Genos-m-derived genome-informed species representations in-corporated into a human microbiome self-supervised learning model improved colorectal cancer (CRC)-control classification over conventional species-abundance random forest models. Genos-m also generated stable sample-level embeddings from as few as 10,000 metagenomic reads, retaining gut microbial community structure that distinguished geographic origin and aligned with enterotypes defined from full-depth taxonomic profiles. Together, these results support Genos-m as a reusable representation model for microbial genomes and metagenomes, with conclusions bounded by the reported datasets, task definitions and evaluation protocols. Genos-m model weights, inference code, and usage documentation are publicly available on GitHub (https://github.com/BGI-HangzhouAI/Genos-m) and Hugging-Face (https://huggingface.co/BGI-HangzhouAI/Genos-m).

bioinformatics↗

bioRxiv2026

Effects of Phthalate Metabolite Mixture Exposure on Mouse Oocyte Development

Phthalates are pervasive endocrine-disrupting chemicals widely used in consumer products. The wide use of many phthalates results in chronic human exposure to complex mixtures rather than single compounds. Despite extensive studies on individual compounds, the combined effects of phthalate metabolites on oogenesis remain poorly understood. Here, we developed a precise microinjection-based single-oocyte toxicological assay to examine the impact of a defined phthalate metabolite mixture on meiotic progression. Phthalate mixture exposure markedly impaired oocyte maturation, as most oocytes failed to extrude the first polar body. Mechanistic analyses revealed severe meiotic defects, including disrupted spindle morphology, chromosome misalignment, disorganized actin cytoskeleton, and impaired mitochondrial function, accompanied by excessive reactive oxygen species (ROS) accumulation and DNA damage. Single-cell transcriptomic profiling further identified differentially expressed genes enriched in biological processes related to exocytosis, secretory pathway regulation, and cytoskeletal organization, as well as in MAPK, JAK-STAT, cGMP-PKG, and GnRH signaling pathways that are essential for follicular development and oocyte maturation. Together, these findings demonstrate that combined phthalate exposure directly compromises female gamete quality and underscore the importance of evaluating mixture effects when assessing risks to womens reproductive health.

cell biology↗

bioRxiv2026

Beclin-1 restrains aldosterone signaling via autophagic degradation of the mineralocorticoid receptor to protect against cardiovascular injury

Cells deploy adaptive programs to maintain homeostasis under stress, yet mechanisms counteracting damage triggered by transmembrane signaling remain poorly defined. Using a hyperaldosteronism model, we examined how autophagy regulates aldosterone-mediated mineralocorticoid receptor (MR) activation. In human umbilical vein endothelial cells (HUVECs), aldosterone induced autophagy, as evidenced by elevated Beclin-1, an increased LC3-II/LC3-I ratio, and reduced SQSTM1/p62. Aldosterone also promoted MR translocation from the cytosol to the nucleus. Co-immunoprecipitation and immunofluorescence revealed direct interaction and colocalization between MR and Beclin-1, as well as enhanced MR-lysosome association. Domain mapping showed that the Beclin-1 middle domain (161-241 AA) binds the MR C-terminal region (601-984 AA). Bioinformatic prediction and ChIP-qPCR confirmed that MR occupies the promoters of IL-1{beta}, IL-6, and TNF- upon aldosterone stimulation. Beclin-1 overexpression attenuated MR nuclear translocation, promoter binding, and inflammatory cytokine expression, whereas Beclin-1 knockdown reversed these effects. In vivo, aldosterone-infused Beclin-1 transgenic (Becn1-tg) mice exhibited lower blood pressure, reduced aortic medial thickening, and attenuated cardiac hypertrophy relative to wild-type controls, with no difference in body weight. Our findings identify Beclin-1 as a critical negative regulator of aldosterone signaling through an autophagy-dependent negative feedback loop. By interacting with MR and directing it toward lysosomal sequestration, Beclin-1 limits MR nuclear translocation and transcriptional activity, thereby mitigating aldosterone-induced vascular inflammation and cardiovascular injury. HighlightsAldosterone activates autophagy and promotes MR-Beclin-1 interaction in HUVECs Beclin-1 binds the C-terminal MR domain and directs MR to lysosomal degradation Beclin-1 overexpression suppresses MR nuclear translocation and cytokine gene activation Beclin-1 transgenic mice are protected from aldosterone-induced cardiovascular injury

molecular biology↗

bioRxiv2026

BioRAG-DRAG: A Multimodal Biological Retrieval Layer for Local-First Biomedical Agents

Biomedical agents need reliable access to heterogeneous evidence: literature text, gene and pathway records, protein sequences, DNA/cDNA sequences, and structured biological relations. Classical sequence tools such as BLAST remain the right choice for alignment-grounded verification, but they are not a unified context interface for large language model agents. We present BioRAG-DRAG, a local-first multimodal retrieval layer that combines pluggable neural sequence-text retrieval, BLAST verification, and graph-based evidence packaging. Specialized encoders such as ESM-2 can serve protein partitions, while OmniGene CPT provides a unified biological-language backbone for mixed sequence/text and agent-facing use; BLAST reranks or verifies sequence candidates; and DRAG graphs expose typed, traceable paths for downstream agents. We introduce BioRAG-Standard v0, a partitioned corpus/library with 257,886 retrievable records and an initial annotation layer for engineering evaluation built from Open-Rosalind Standard biomedical records and sequence-window extensions. On an in-index sequence-window stress test, BLAST nearly saturates biological matching, while vector retrieval recovers substantial but lower biological match rates. On held-out parent-fragment controls, public protein encoders outperform the current OmniGene protein-window embedding, while DNA/cDNA dense retrieval remains weak even with off-the-shelf Nucleotide Transformer pooling; this supports a model-agnostic BioRAG design rather than a claim that one unified generator backbone is the best sequence-search encoder. Indexed Chroma lookup over Standard text and 100k sequence-window collections adds only small lookup overhead after query embedding; this does not measure end-to-end instant latency. Finally, exploratory sequence DRAG traces show inspectable biological neighborhoods, including immunoglobulin-family and gene-symbol modules, with initial graph controls indicating non-random but partly sequence-similarity-driven structure. These results support a bounded architecture: vector retrieval supplies unified candidate context, while BLAST and DRAG provide biological verification and evidence attribution.

bioinformatics↗

bioRxiv2026

Combinatorial transcription factor interactions drive modular gene regulatory networks

Transcription factors (TFs) cooperatively drive gene regulatory networks (GRNs) to establish transcriptional states. Forced induction of TFs in combination can reprogram cell state by supplanting existing GRNs. Thus, TFs and GRNs are the building blocks to engineering transcriptional state. However, one key challenge is that the relationship between TF combinations and GRNs remains largely uncharacterized and difficult to accurately predict. Here, we apply single-cell overexpression screens to map the combinatorial activities of [~]100 TFs to gene expression states. Our analysis identifies diverse TF combinations driving cell-type specific regulatory programs. Notably, different TF combinations induce shared gene sets with cell-type specific functions, suggesting a modular regulatory architecture of the transcriptome. Furthermore, we define pairwise TF interactions and show that cooperative interactions improve transcriptional reprogramming. Finally, we developed tools to predict combinatorial TF phenotypes. These findings improve our understanding of cell state and how to manipulate it for biomedical applications. HIGHLIGHTSO_LICombinatorial over-expression screens for [~]100 transcription factors (TFs). C_LIO_LIDiverse TF combinations drive cell-type specific regulatory programs. C_LIO_LITF regulatory networks reveal a modular regulatory architecture of the transcriptome. C_LIO_LITF-TF interactions and predictive models enhance reprogramming cocktails. C_LI

genomics↗

bioRxiv2026

Gut Microbiota-derived Acetate Safeguards the Colonic Epithelial Acetyl-CoA Reserve to Avert Colonic Senescence

Aging is a major driver of tissue senescence, but little is known about the development of age-unrelated tissue senescence. Here we show that nucleocytosolic acetyl-CoA deficiency in colon epithelial cells, caused by Acly ablation and bacterial depletion, triggers age-independent, p53-dependent colonic senescence leading to severe systemic inflammation. Acetate supplementation, acetate-producing bacterial transplantation, targeted depletion of senescent cells or treatment with lysine deacetylase inhibitors blocks colonic senescence and inflammatory injury. Spontaneous colonic senescence develops following simultaneous deletion of epithelial Acly and Acss2, confirming that microbe-derived acetate maintains the epithelial acetyl-CoA pool via ACSS2 to avert colonic senescence. Mechanistically, acetyl-CoA deficiency deprives a cohort of mitochondrial and nuclear proteins of acetylation, leading to increased oxidative stress and DNA repair stress that trigger cellular senescence. Among these proteins, ATP5F1A-K161 acetylation and H4-K5/8 acetylation are required to protect colonic epithelial cells from developing senescence. These observations unveil a previously unknown mechanism that governs colonic senescence.

physiology↗

bioRxiv2026

Melanin regulates mitochondrial dynamics, metabolism and inflammatory signaling to protect the retina

Albino individuals are clinically recognized to exhibit heightened susceptibility to light-induced retinal injury, yet the cellular and metabolic mechanisms underlying this vulnerability remain poorly defined. Here, we investigated whether retinal pigment epithelium (RPE) pigmentation governs mitochondrial structure, metabolism, and inflammatory responses that ultimately determine retinal resilience to blue light stress. Using pigmented (C57BL/6J) and albino (Balb/c) mice, we demonstrate that albino animals exhibit markedly increased retinal phototoxicity following blue light exposure, manifested by fundus lesions, outer nuclear layer (ONL) disruption, and structural degeneration evident by OCT. Primary RPE cultures derived from albino mice exhibited profound difference in mitochondrial morphology, characterized by increased mitochondrial number, reduced size, and enhanced fragmentation, accompanied by elevated mitochondrial DNA copy number. These structural changes correlated with transcriptional skewing toward mitochondrial fission (increased Drp1) and suppression of mitochondrial fusion (Mfn1, Mfn2, OPA1). Functionally, albino and depigmented RPE displayed impaired oxidative phosphorylation, reduced ATP production, and diminished reliance on mitochondrial pyruvate carrier (MPC)-dependent metabolism. In parallel, albino RPE demonstrated cell-cycle accumulation at G2/M and heightened basal and blue light-induced secretion of pro-inflammatory cytokines, particularly IFN-{beta}1, IL-6, and TNF-. Importantly, exogenous melanin supplementation partially restored mitochondrial fusion gene expression, pyruvate-dependent respiration, and inflammatory restraint. Together, these findings identify melanin as a critical regulator of RPE mitochondrial architecture, metabolic substrate utilization, and inflammatory signaling, providing a mechanistic framework to explain enhanced photo-vulnerability in the albino retina. These insights establish pigmentation-dependent mitochondrial metabolism as a determinant of retinal resilience and suggest mitochondrial bioenergetics as a therapeutic target.

cell biology↗

bioRxiv2026

OmniGene-4: A Unified Bio-Language MoE Model with Router-Level Interpretability

How do multi-modal large language models that jointly process natural language and biological sequences (DNA, protein, structural alphabets) actually answer biological questions, especially sequence-grounded questions whose answer depends on residue-level patterns rather than literature recall? We introduce OmniGene-4, a unified bio-language Mixture-of-Experts foundation model on Gemma-4-26B-A4B (128 experts/layer, top-8 routing), and use its discrete router state to dissect this question. By hooking every router across eight task families, we provide the first router-level decomposition for a biological MoE: continued pretraining (CPT) accounts for 96% of cross-task expert differentiation and supervised fine-tuning (SFT) for 4%, reshaping middle and output layers respectively. Within the protein-homology task family, per-pair routing divergence stays below 0.04 (vs 0.23 cross-task), implying that sequence-grounded decisions occur inside expert computation rather than at the gate -- the gate selects the modality, the experts compute the answer. The pipeline yields strong benchmarks: remote-homology 82.60% (vs ESM-2 3B, MMseqs2, DIAMOND by 28-31 pp); standard homology 99.40%; BixBench (general biological-knowledge) 93.66%. A dual-head architecture adds per-residue 3Di/DSSP classifiers (78.6%/100%). To probe whether the discovered transfer mechanism is robust under modality scaling, we further extend the model to OmniGene-4-MM, adding four vision modalities (chemical-structure images, medical/pathology imagery, charts) via a vision tower and a three-stage LoRA pipeline at 1.5 GPU-days total. The multi-modal model preserves the homology capability (85% standard, 69.5% remote) and acquires chemist-readable structure understanding (96% on Vis-CheBI20 functional-group captioning) while consuming roughly four orders of magnitude less compute than recent specialized MoE bio-models. The work characterizes how multi-modal bio-foundation models acquire, route, and preserve sequence-aware capability -- central to the next generation of scientific large language models. O_TEXTBOXThe bigger picture. Modern AI models that read both human language and biological sequences (DNA, proteins) often behave like black boxes: we see their answers but not the inner mechanism that produced them. This matters for biology, where a wrong but confidently worded answer can mislead an experiment that costs months and tens of thousands of dollars. We use Mixture-of-Experts architecture -- a transformer where each token is routed to a small subset of 128 specialized sub-networks at every layer -- to make this internal mechanism legible. By logging which experts are activated for each input, we show that adding biology-specific pre-training to a general-purpose language model causes the routing to spontaneously partition the network into modality-specific sub-networks, and that the same partitioning re-emerges when we further extend the model to also process molecular images, medical images, and chart images. The same model achieves protein-homology accuracy that surpasses classical sequence-alignment tools (MMseqs2, DIAMOND) and recent protein language models (ESM-2 3B), at roughly four orders of magnitude less compute than recent specialized MoE bio-models. The work is a step toward foundation models for biology that are simultaneously broad in modality coverage, mechanistically transparent, and economically reproducible by groups outside the largest industrial labs. C_TEXTBOX

bioinformatics↗

bioRxiv2026

Covalent tumor anchoring spatially orchestrates antitumor immunity

Protein immunotherapies can elicit potent tumor rejection, but reversible target engagement, incomplete tumor retention, and systemic leakage often erode spatial control. Here, we develop covalently anchored tumor immunotherapeutic proteins (CATIPs), a modular platform that uses proximity-enabled covalent chemistry to immobilize immune cues on tumor-cell surfaces after intratumoral administration. CATIPs combine tumor-targeting nanobodies with payloads for T cell engagement, co-stimulation, and cytokine support. In human PBMC-reconstituted NSG mice, CATIPs completely eradicated treated EGFR-positive tumors, outperforming matched non-covalent proteins while limiting redistribution, systemic T cell activation, cytokine release, xGVHD-associated morbidity, and on-target, off tumor toxicity. In immunocompetent melanoma models, CATIPs remodeled the tumor microenvironment, expanded antigen-specific CD8+ T cells, induced antigen-restricted abscopal control, and generated durable protection against local and metastatic rechallenge. CATIP-engineered tumor cells further functioned as whole-cell vaccines. Thus, covalent tumor anchoring converts local protein delivery into tumor-surface immune programming, enabling systemic, tumor-specific, durable antitumor immunity while limiting systemic immunopathology.

synthetic biology↗

bioRxiv2026

A graph-based pangenome reveals the genetic basis of climate-resilient and horticultural traits in pear

Climate change poses an increasing threat to the cultivation of deciduous fruit trees, placing greater demands on modern pear breeding. Using pear germplasm adapted to diverse environments, we assembled 11 chromosome-level genomes. In combination with 13 publicly accessible pear genomes, we analyzed presence-absence variations (PAVs) and constructed a graph-based pangenome for pear. By performing a PAV-eQTL analysis of the fruit of 123 pear accessions, we identified PAVs significantly associated with expression levels of genes that may be involved in regulating agronomic traits. Population analysis of 268 pear accessions revealed two stop-gained variants in DAM1 of independent origin, which may function in advancing the blooming date and reducing the chilling requirement. We detected complex PAVs at the NOR1 locus, including two copy-number variations and one deletion. These PAVs contributed to the rapid diversification of the NOR1 locus and the fruit development period through regulating ARF5 and other ripening-related genes. We revealed the selection history of the NOR1 locus and developed novel pear individuals that accumulated alleles for low chilling requirement, early blooming date, and short fruit development period. The results provide valuable resources for pear genomics research and offer a guideline for breeding modern pears with climate resilience.

plant biology↗

bioRxiv2026

Temporal-deviation-driven community detection uncovers early-warning signals for critical transitions in complex diseases

Early detection of critical transitions in complex diseases is crucial for timely clinical intervention. However, as patients often provide only a single snapshot, identifying sample-specific early-warning signals (EWS) from a dynamical evolution perspective remains challenging, coupled with high-dimensional noise amplification. Here, we present TD-COM, a framework for detecting personalized EWS of critical transitions via single-sample community detection. By constructing a temporal perturbation map STDN, TD-COM captures latent dynamical perturbations inferred from static individual profiles. Synergizing these temporaldeviation signals with static topological features, TD-COM implements a multilevel node filtering strategy during community detection, effectively suppressing single-sample noise. Validated on hour-scale, multi-year, and multi-decade transcriptomic data, TD-COM robustly detects critical states preceding clinical deterioration and uncovers their underlying molecular mechanisms. Comparative experiments demonstrate that TD-COM outperforms existing methods in accuracy and topological robustness. Thus, TD-COM provides a generalizable framework for personalized early warning of complex diseases, particularly when longitudinal sampling is infeasible.

bioinformatics↗

bioRxiv2026

NO modulates human airway smooth muscle function by altering glucose-6-phosphate dehydrogenase effects on sGC function in asthma

Since NO can modulate mesenchymal cell function, we posit that NO can modulate gene expression associated with excitation-contraction coupling. Our study shows that treating asthma-derived HASMCs with a low dose of NO plus sGC stimulator BAY-41, in most cases sensitized smooth muscle sGC towards activation via an elevated sGC heterodimer and in some cases also improved sGC{beta}1, catalase, Cyb5r3 or Trx1 expression (n=24 non-asthma and n=25 asthma). Interestingly we found that majority of asthma HASMCs showed a marked downregulation of G6PD expression inducing a low GSH/GSSG ratio in asthma, and these findings were replicated in murine lungs of allergic asthma (OVA and CFA/HDM). Studies with HEK/COS-7 cells showed G6PD synergizing with hsp90 in enabling sGC heme-maturation. G6PD overexpression in HASMCs enhanced the sGC heterodimerization while silencing of endogenous G6PD abrogated it. Complementation of these cellular results with whole animal models of G6PD deficiency or overexpression provided verification to our findings. Mouse lung tissue from the humanized variant of G6PD deficiency, V68M (G6PD A-deficiency) showed significant downregulation in the sGC heterodimer, with a concomitant reduction in its NO heme-dependent activity, thereby showing that G6PD deficiency lowers sGC heme. Conversely, G6PD overexpressing mouse lung tissue displayed an elevated sGC heterodimer and also showed a robust G6PD-sGC{beta}1 interaction, suggesting G6PD to be involved in the heme-maturation of sGC{beta}1. While G6PD maintains the cell redox by generating NADPH, its new role in regulating sGC maturation links sGC dysfunction in asthma to G6PD deficiency and may potentially uncover new targets for asthma treatment.

cell biology↗

bioRxiv2026

Sex-Dependent Effects of Glutamatergic Disruption on Dopaminergic Neuron Subtype Vulnerable in Parkinson's Disease

Aldehyde dehydrogenase 1A1-positive (ALDH1A1+) dopaminergic neurons (DANs) are preferentially vulnerable in Parkinsons disease (PD), yet how their activity is modulated by presynaptic inputs remains poorly defined. Here we investigated the role of glutamatergic input by conditionally deleting Grin1, which encodes a critical NMDA receptor (NMDAR) subunit, in ALDH1A1+ DANs. Grin1 conditional knockout (cKO) mice displayed normal locomotion and motor learning; however, females exhibited enhanced operant reward acquisition and excessive feeding with transient weight gain following food restriction. To determine regional contributions, Grin1 was selectively knocked down in ALDH1A1+ DANs of either the ventral tegmental area (VTA) or substantia nigra pars compacta (SNc). VTA-specific knockdown in females was sufficient to reproduce the post-restriction feeding and weight gain phenotype. Bulk whole-brain mRNA sequencing revealed pronounced sex-dependent transcriptional changes, primarily in female Grin1 cKO mice after food restriction. Many differentially expressed genes were associated with mitochondrial function, energy metabolism, and synaptic signaling. Together, these findings reveal a sex-specific role for NMDAR-mediated glutamatergic input to ALDH1A1+ VTA DANs in regulating feeding behavior, providing mechanistic insight into how dysfunction of this vulnerable subpopulation may contribute to PD-associated compulsive eating disorders. Key FindingDisrupted NMDA receptor-mediated glutamatergic input to ALDH1A1+ DANs drives sex-specific feeding abnormalities relevant to PD-associated compulsive eating disorders.

neuroscience↗

bioRxiv2026

Open-Rosalind: Tool-First Biomedical LLM Agents with Process-Aware Benchmarking

Large language models are increasingly used as scientific agents, yet the flexibility that benefits general-purpose agents can conflict with the accountability required in biomedical research. We study whether biomedical agents can be organized around auditable constraints rather than unconstrained autonomy. We present Open-Rosalind, a tool-first bio-agent system designed around four operational principles: evidence-grounded outputs, trace completeness, workflow-constrained execution, and explicit tool mediation for factual claims. To evaluate these principles, we introduce Open-Rosalind BioBench, a process-aware benchmark that measures not only task accuracy but also tool correctness, citation presence, trace completeness, and failure rate. On a strict in-house benchmark, the reference pipeline achieves 81.4% accuracy with complete execution traces. In multi-model ablations and paired replications, removing tools reduces accuracy by 19.3 to 26.4 percentage points, indicating that tool-first execution is the strongest and most stable contributor to performance. Constrained workflows also reduce lower-tail failures for models that are weak at free-form tool use. However, an author-independent 30-task hold-out initially revealed severe external-validity collapse on the deployment model. After diagnosing five routing and normalization failures and applying targeted fixes, hold-out accuracy improved from 17.8% to 53.3%, and the most concerning negative comparison against a no_tool baseline disappeared. These results position Open-Rosalind as a biomedical-agent study with an explicit external-validity audit, rather than as a claim that protocol constraints alone guarantee superior performance.

bioinformatics↗

bioRxiv2026

Parallel Activation and Interference CRISPR (PAIR) with Sequencing Uncovers DNA Repair Networks Guiding Precision Cell Engineering

The dynamic balance of cellular homeostasis is often maintained by opposing regulatory pathways, yet most genetic screens interrogate them in one direction and therefore miss the bidirectional gene-gene interactions that shape complex phenotypes such as DNA damage response (DDR). Here, we present PAIR (Parallel Activation and Interference CRISPR), a bidirectional perturbation platform that enables simultaneous activation and suppression of distinct genes within the same cell using CRISPR activation (CRISPRa) and Cas13d RNA knockdown. Applying PAIR to the CRISPR/Cas9 induced DSB repair screen, we mapped gene-gene interactions across competing repair branches and identify synergistic perturbations, including NBN activation combined with suppression of end-joining factors, that shift repair outcomes toward homology-directed repair (HDR) and improve the precision of CRISPR-based gene editing. Using coupled PAIR with single-cell transcriptomic, we further demonstrated that NBN activation induces inflammatory and interferon programs, whereas co-suppression of end-joining factors buffers this response, revealing transcriptional states missed by conventional unidirectional perturbations. To translate these findings into non-viral chimeric antigen receptor (CAR) T cell engineering, we developed an mRNA-based strategy for parallel overexpression and knockdown of NBN-anchored DDR effectors in primary T cells, priming the T cells into a transient HDR-favored state that enhances the efficiency of CAR knock-in on the TRAC locus. Together, the PAIR system provides a general framework for studying opposing regulatory networks, uncovering hidden cell states, and guiding cell-state engineering through bidirectional perturbation.

bioengineering↗

bioRxiv2026

Neuromodulation of Foxp2+ hypothalamic neurons induces therapeutic hypothermia

The inability to selectively trigger therapeutic hypothermia independent of environmental cooling has hindered causal analysis of its broader physiological benefits. Here, we employed P57, a natural product that pharmacologically induces therapeutic hypothermia circumventing external cold stress, to identify the neuronal substrates underlying hypothermia-induced antitumor effects. Integrating functional ultrasound imaging, activity-dependent neuronal labeling, and single-nucleus RNA sequencing, we identify a previously unrecognized population of Foxp2 neurons in the medial preoptic area (MPA) that are selectively activated by P57. Inhibition of MPA Foxp2+ neurons abolishes P57-induced hypothermia, whereas the paraventricular hypothalamus (PVH) serves as a critical downstream node. Importantly, targeted activation of MPA Foxp2 neurons is sufficient to induce sustained hypothermia, suppress systemic metabolism, and inhibit tumor growth, providing causal evidence that neuronally driven reductions in core body temperature can exert antitumor effects. Together, these findings establish MPA Foxp2 neurons as a controllable node for therapeutic hypothermia induction and demonstrate that neuromodulation of defined neuronal populations can achieve therapeutic benefit by directly controlling physiological states.

neuroscience↗

bioRxiv2026

Multimodal profiling reveals the centromedian nucleus of thalamus as a dynamic hub orchestrating staged consciousness recovery

Consciousness recovery from general anesthesia represents a fundamental brain state transition. The absence of objective, continuous metrics to monitor the shift from unconsciousness to full awareness has left its dynamic, multidimensional nature unresolved, fracturing consciousness theoretical framework and posing clinical risks due to unreliable assessment of emergence. To address this, we developed a multiscale framework integrating AI-driven behavioral analysis, continuous neuro-physiology recording, and whole-brain functional imaging. This decodes recovery into three hierarchical stages, reflex restitution, level restoration, and content re-establishment, each with unique multimodal signatures. We identified heart rate stabilization as a potential noninvasive biomarker for the restoration of consciousness level. Crucially, the centromedian nucleus of thalamus acts as a dynamic hub, actively orchestrating staged whole-brain reconfiguration via stage-dependent network routing. These findings resolve the temporal orchestration of consciousness recovery, ground disparate theories within a unified mechanistic sequence, and open transformative paths for real-time monitoring and neuromodulation in anesthesiology. HighlightsA multimodal framework decodes consciousness recovery into three hierarchical stages Recovery stages span from reflex restitution, to level restoration, and finally to content re-establishment Heart rate stabilization serves as a noninvasive biomarker for consciousness level restoration The CM paces this staged reconfiguration as a dynamic network hub In BriefZhang et al. develop a multiscale framework that delineates consciousness recovery from anesthesia as a three-stage process, identifies heart rate stabilization as a biomarker for the restoration of consciousness level, and reveals the CM as a dynamic hub orchestrating this progression. These findings clarify the temporal orchestration of consciousness recovery, pave the way for real-time monitoring and neuromodulation in anesthesiology and other fields.

neuroscience↗