Biology subjects

Huang, J.

Publications and source records attributed to Huang, J..

At least 127 records · Page 7Linked to original sources

bioRxiv2025

Genetic Associations with Longevity in a Calabrian Cohort: A Genome-Wide Study

Human longevity is a complex trait shaped by genetic background and population-specific factors. Calabria, a region in Southern Italy with a high prevalence of centenarians and relative genetic isolation, is a valuable model for investigating the genetic architecture of extreme survival. Here, we performed a genome-wide association study of longevity in 705 Calabrian individuals, comparing long-lived subjects to younger controls using a mixed-model approach that accounts for relatedness and population structure. We identified 267 candidate longevity-associated variants, including 23 suggestive genomic risk loci, of which one reached genome-wide significance. Although most loci did not replicate in external datasets, one intronic variant regulating proteasome-related gene expression was confirmed by meta-analysis. Gene- and pathway-based analyses highlighted biological processes central to aging, including proteostasis, DNA repair, telomere maintenance, apoptosis, insulin signaling, inflammation, and cancer-related pathways. Notably, established longevity loci such as APOE and FOXO3 were not associated, underscoring population-specific genetic effects. Overall, our findings suggest that longevity in the Calabrian population arises from a combination of unique genetic influences and conserved aging-related mechanisms, providing new insights into the molecular basis of human lifespan extension.

genomics↗

bioRxiv2025

IGHV3-53 antibody abundance drives divergent SARS-CoV-2 immune imprinting

The mechanisms driving divergent SARS-CoV-2 immune imprinting in populations with distinct SARS-CoV-2 exposure histories remain unclear. However, conventional wild-type mouse models fail to recapitulate severe imprinting, even after potent ancestral-strain mRNA immunization, hindering imprinting-related mechanistic investigation and vaccine evaluation. Here, we surprisingly found that V(D)J-humanized mice could faithfully recapitulate human severe imprinting phenotypes. Comprehensive antibody repertoire analysis and deep mutational scanning-based epitope mapping of 583 monoclonal antibodies from these models revealed that the pre-existing IGHV3-53/66 antibody abundance determine imprinting severity through antibody-mediated masking of Omicron-specific epitopes. Both passive IGHV3-53/66 antibody transfer and IGHV3-53 knock-in were sufficient to induce severe imprinting in wild-type mice. Together, these findings demonstrate that the V(D)J germline repertoire--even a single germline-encoded antibody response--can profoundly shape humoral imprinting severity. Accordingly, we established an IGHV3-53 knock-in mouse model that accurately recapitulates the human antibody landscape, providing a valuable tool for guiding future COVID-19 vaccine updates.

immunology↗

bioRxiv2025

The Mycobacterium tuberculosis Ku C-terminus orchestrates LigD activity through domain-specific interactions

Bacterial non-homologous end joining (NHEJ) is a DNA double-strand break (DSB) repair pathway that relies on the Ku-LigD complex to alleviate genomic instability. The Mycobacterium tuberculosis Ku C-terminus has been highlighted for its role in LigD recruitment to DNA DSBs and stimulation of ligase activity. However, it remains unclear how the Ku C-terminus interacts with and potentially influences other LigD activities. Here, we combine NMR spectroscopy, structural modelling and mutational analysis to define the interaction interface between the Ku C-terminus and LigD. We identify critical residues in Ku (E246, V248, S258, K260, and N266) and the LigD polymerase (D162, V194, R198) and ligase (D522, K579, L580) domains that mediate this interaction. Functional assays reveal that Ku stimulates LigD ligase activity through contacts with both polymerase and ligase domains, while Ku attenuates template-dependent polymerase activity, contrasting previous studies with Pseudomonas aeruginosa homologs. Disrupting the Ku-LigD interface, either through Ku or LigD mutations, abolishes ligase stimulation and restores polymerase activity, highlighting a dual regulatory mechanism. Our data supports a model where Kus C-terminal region forms a bipartite interface with LigD to balance repair activity. These findings provide mechanistic insight into the Ku-LigD repair mechanism and uncovers species-specific differences in bacterial NHEJ. Graphical Abstract O_FIG O_LINKSMALLFIG WIDTH=200 HEIGHT=81 SRC="FIGDIR/small/693941v2_ufig1.gif" ALT="Figure 1"> View larger version (20K): org.highwire.dtl.DTLVardef@802e3org.highwire.dtl.DTLVardef@1f5ac14org.highwire.dtl.DTLVardef@1f46677org.highwire.dtl.DTLVardef@7312b5_HPS_FORMAT_FIGEXP M_FIG C_FIG

biochemistry↗

bioRxiv2025

Fast Loading and Transformation of Large-Volume Bioimaging Data Consistently Approaching the I/O Peak Speed

High-resolution large-volume biological imaging techniques are now widely used in biological research, but inefficiencies in data processing and visualization persist due to bottlenecks in loading/saving pipelines and limitations of conventional pyramid formats. To address these issues, we have decoupled data loading process into reading from disk and in-memory decompression, while data saving has been separated into in-memory compression and writing to disk. This approach allows us to dedicate more CPU cores to decompression and compression tasks, optimizing resource allocation and maximizing disk I/O throughput. As a result, both loading and saving processes operate near the disks peak data transfer capacity. Further, we integrated a video stream-inspired format into the pyramid structure, allowing direct access to regions of interest (ROIs) without loading extraneous data into RAM, reducing memory overhead. Compared to the latest methods, our approach improves pyramid data transformation efficiency by at least 6x and dramatically accelerates downstream tasks like visualization and deep learning inference. TeaserNovel bioimaging data processing pipeline achieves near-peak I/O throughput via decoupled architecture and video-formatting.

bioinformatics↗

bioRxiv2025

Armoring STEAP1 CAR T cells with IL-18 potentiates antitumor activity in Ewing sarcoma

BackgroundEwing sarcoma (EwS) is a highly aggressive cancer driven by the EWS::FLI1 fusion oncoprotein affecting children, adolescents, and young adults. Six transmembrane epithelial antigen 1 (STEAP1) is a cell surface antigen transcriptionally controlled by EWS::FLI1 that is broadly expressed in EwS, positioning it as a rational immunotherapy target. However, translating CAR T therapy to solid tumors requires overcoming barriers to potency while maintaining safety. MethodsAnalyses of transcriptome and proteome data were performed to evaluate the effects of EWS::FLI1 perturbation on STEAP1 expression at the transcript and protein levels in EwS models. STEAP1 expression was validated in EwS patient tissues by immunohistochemistry. Second-generation STEAP1-BB{zeta} CAR T cells were tested in orthotopic and disseminated EwS xenograft models. To enhance antitumor activity, an IL-18-armored STEAP1 CAR was engineered. Dose-dependent therapeutic efficacy and safety were evaluated through measurement of tumor burden, survival, and observation for gross toxicities. ResultsSTEAP1 was expressed in [~]97% of primary EwS tumors and directly associated with EWS::FLI1 fusion protein expression in EwS cell lines. In orthotopic EwS models, STEAP1 CAR T cells induced complete tumor regression at 5 x 106 cells. In disseminated disease models, responses were dose-dependent with no evidence of antigen loss. Notably, IL-18 armored STEAP1 CAR T cells achieved complete responses in [~]80% of mice at a reduced dose of 106 cells without overt toxicity. ConclusionsThese data establish STEAP1 as a clinically relevant and highly expressed target in EwS and demonstrate that IL-18 armoring significantly improves CAR T cell efficacy by enhancing potency evident through antitumor activity at reduced cell dose. STEAP1 CAR T cells are currently under evaluation in a first-in-human phase 1/2 dose-escalation clinical trial for metastatic castration-resistant prostate cancer (NCT06236139) and these studies support future clinical translation of STEAP1 CAR T cell therapy for relapsed/refractory EwS.

cancer biology↗

bioRxiv2025

Super-CUT&Tag: A Sensitive, Spatially Resolved Approach for Epigenomic Profiling in Tissue Sections

Spatial profiling of protein-DNA interactions is essential for understanding gene regulatory processes, but current methods suffer from low sensitivity, limited resolution, and complex workflows. Here we introduce Super-CUT&Tag (Solid-phase interface embedded with an Ultra-dense DNA barcode array to Profile Epigenomic landscapes for spatially Resolved CUT&Tag), a robust spatial epigenomic method that directly transfers in situ protein-DNA interactions from tissue sections onto a pre-fabricated spatially barcoded array. Incorporating 3D polyamidoamine (PAMAM) dendrimers markedly improves chromatin capture efficiency, achieving several-fold higher sensitivity compared to current spatial CUT&Tag approaches. Using Super-CUT&Tag, we profile H3K27ac spatial landscape and uncover previously hidden active-chromatin heterogeneity across mouse embryonic tissues and stages. Integration with spatial transcriptomics further reveals spatiotemporal enhancer regulation dynamics of Neurod2 during corticogenesis. By overcoming key bottlenecks in sensitivity, resolution and usability, Super-CUT&Tag establishes a powerful platform for spatial epigenomic profiling in complex tissues.

genetics↗

bioRxiv2025

scProtoTransformer: Scalable Reference Mapping Across Molecules, Cells and Donors

The rapid accumulation of single-cell data has made it possible to comprehensively characterize biological systems at molecular, cellular, and donor levels. However, scalable reference mapping across different resolutions remains a major challenge in current research. Here, we propose scProtoTransformer, a prototype-based Transformer architecture designed to achieve scalable reference mapping across molecular, cell, and donor levels. scProtoTransformer introduces a knowledge-guided prototype tokenizer that projects gene expression into biologically interpretable pathway prototypes, effectively reducing numerical batch effects while preserving biological semantic patterns. Furthermore, by leveraging knowledge distilled from the foundation model and a dynamic supervised fine-tuning strategy, scProtoTransformer achieves robust biological representations with reduced pre-training requirements. Benchmark experiments across molecular, cell, and donor-level reference mapping demonstrate that scProtoTransformer delivers competitive or even superior performance compared with state-of-the-art approaches, while providing interpretability through biologically prototypes. Together, these results establish scProtoTransformer as a unified framework for scalable reference mapping, laying the foundation for systematic understanding from genes to individuals.

bioinformatics↗

bioRxiv2025

A tailored in vivo CRISPR screen identifies BAP1 as a potent tumor suppressor of soft tissue sarcoma

Undifferentiated pleomorphic sarcoma (UPS) is one of the most common soft tissue sarcomas (STS) in adults. Despite decades of research, therapeutic advancements for STS, including UPS, have remained limited. The genetic complexity of UPS, characterized by the absence of recurrent driver oncogene mutations, has hindered the development of effective targeted therapies beyond conventional chemotherapy and immunotherapy. To address this challenge, we conducted a customized in vivo CRISPR/Cas9 screen in mice to systematically identify potential tumor suppressors involved in UPS development. Our screen revealed BRCA1-associated protein 1 (Bap1) as a potent tumor suppressor in STS. Using total RNA sequencing, multiplex immunohistochemistry, and flow cytometry, we found that Bap1-deficient mouse sarcomas exhibit significant immune suppression. Further analysis indicated that polo-like kinase 1 (Plk1) is essential for the survival of Bap1-deficient sarcomas. Treatment with volasertib, a Plk1 inhibitor, markedly inhibited tumor growth in both syngeneic and spontaneous mouse models of Bap1-loss sarcoma. In conclusion, our findings suggest that PLK1 inhibition, or combined with immunotherapy, may represent a promising targeted therapeutic strategy for tumors lacking BAP1.

cancer biology↗

bioRxiv2025

Alphaviral nonstructural protein-host RBP co-condensation as a mechanism to sustain virus replication

It has long been recognized that the intracellular replication of alphaviruses critically relies on several key host RBPs, including G3BP1/2 and FXR1/FXR2/FMR1, but how these RBPs modulate alphaviral replication and whether it would be possible to target these RBPs for antiviral treatment are less explored. Here, using SFV as a model, we report that SFV nsP3 exploits G3BP for its condensation and transforms antiviral stress granules into proviral nsP3-G3BP co-condensates. The gel-like co-condensates of nsP3 and G3BP enrich and protect viral genomic RNAs from host RNase degradation and serve as viral translational hubs to promote viral replication. The mode of nsP3-RBP co-condensation is prevalent across alphaviruses, and disruption of nsP3 condensates is an efficient antiviral approach. Thus, these findings uncover a general anti-alphavirus strategy based on the conserved reliance of nsP3-RBP co-condensation.

cell biology↗

bioRxiv2025

Susceptibility to Task-irrelevant Auditory Distractors in Relation to Visual Working Memory in Children With and Without ADHD

ObjectiveAttention-deficit/hyperactivity disorder (ADHD) is a prevalent neurodevelopmental disorder in children and is often associated with increased distractibility. Previous research has reported that susceptibility to task-irrelevant visual distractions is associated with visual working memory (WM) capacity. However, how neural responses to task-irrelevant auditory distractors are related to visual WM capacity in children with and without ADHD remains poorly understood. MethodsIn this study, we collected electroencephalography (EEG) signals from 72 children with ADHD and 88 typically developing (TD) children across two experiments, during which they performed visual tasks of varying difficulty while task-irrelevant auditory distractors were presented. The mismatch negativity (MMN) amplitude was measured as a neural index of susceptibility to auditory distractors, and the K score was assessed as a behavioral indicator of visual WM capacity. ResultsTD children showed opposite patterns across the two experiments: under low audiovisual competition, those with higher visual working memory capacity were more susceptible to task-irrelevant sounds, whereas under high competition, children with lower capacity showed greater susceptibility to distraction. In contrast, children with ADHD showed a stable association, with lower visual WM capacity consistently predicting higher susceptibility to task-irrelevant auditory distractors regardless of competition level. ConclusionsThese findings provide novel evidence that susceptibility to task-irrelevant auditory distraction is closely related to visual WM capacity in both children with and without ADHD, but the patterns of this relationship differ between typical and atypical neurodevelopment and are modulated by audiovisual competition.

neuroscience↗

bioRxiv2025

Developmental Trajectories of Addition and Subtraction in Early School-aged Children

Mastering basic arithmetic lays the foundation for lifelong mathematical achievement, yet how the neural mechanisms underlying addition and subtraction calculations develop during childhood remain unclear. This study investigated the developmental trajectories of addition and subtraction calculations in school-aged children, and explored how arithmetic-specific neural responses in the right parieto-occipital region are related to age-related improvements in behavioral performance. Here, we recorded electroencephalography (EEG) signals from 115 typically developing children divided into three age groups (7-8-, 9-10-, and 11-12-year-olds) during an arithmetic task involving addition and subtraction calculations. We analyzed the P300 component from the original waveforms and the {Delta}P300 component from the difference waveforms, which were generated by subtracting the event-related potentials (ERPs) elicited by the first operand from those elicited by the second operand to isolate arithmetic-specific neural processes. The results revealed that, in both calculations, older children showed better behavioral performance (faster reaction time and larger PI) and greater right parieto-occipital {Delta}P300 amplitudes, with a key developmental stage at 11-12 years. Furthermore, all groups of children exhibited greater right parieto-occipital P300 and {Delta}P300 amplitudes in subtraction, whereas the smaller right parieto-occipital {Delta}P300 amplitudes in addition significantly mediated the relationship between age and PI. These findings reveal distinct patterns for addition and subtraction calculations over development, highlighting the right parieto-occipital {Delta}P300 amplitude as a neural marker of arithmetic performance, particularly for addition calculation.

neuroscience↗

bioRxiv2025

A mouse model of classical trigeminal neuralgia via intradural compression of the trigeminal nerve

IntroductionTrigeminal neuralgia (TN) is a debilitating orofacial pain condition that adversely affects quality of life. Although heterogeneous, the most common form of TN is classical TN, characterized by paroxysmal bouts of pain in response to otherwise innocuous stimuli. It is believed that classical TN results from neurovascular compression of the trigeminal nerve. However, the underlying pathophysiology of TN is not well understood, thus limiting the development of targeted therapies. Current animal models lack translational relevance, particularly in their inability to replicate intradural nerve root compression, a core anatomic component of TN. MethodsWe developed a TN mouse model that achieves intradural nerve root compression via a retro-orbital approach confirmed by anatomic dissection and magnetic resonance imaging. To assess behavioral outcomes, we measured orofacial pain through facial wiping and interaction with a reward stimulus. Pharmacological responsiveness was tested using carbamazepine administration. Mechanistic studies included calcium imaging of trigeminal ganglia (TG), electrophysiologic recordings to measure resting membrane potential and rheobase, and immunohistochemical analysis of the TG. ResultsThe model elicited orofacial neuropathic pain, substantiated by increased facial wiping and reduced interaction with a reward stimulus, behaviors that suggest both spontaneous and evoked pain. Carbamazepine attenuated these behaviors, suggesting pharmacologic relevance to current TN treatment. Calcium imaging showed heightened spontaneous activity in the TG, and electrophysiologic recordings revealed an increased resting membrane potential and a reduced rheobase. Finally, immunohistochemical studies showed infiltration of CD45+ cells, demyelination and an increase in CGRP expression in the TG, supporting the presence of neuroinflammation after nerve root compression. ConclusionThese findings show that our approach replicates the anatomy and clinical presentation of classical TN in humans. This model may represent a new and robust platform for future mechanistic studies of TN and subsequent preclinical evaluation of therapies in mice.

neuroscience↗

bioRxiv2025

The cloacal outgrowth orchestrates co-development of the bladder and umbilical arteries

The urinary bladder is an innovation of eutherian mammals to manage liquid wastes. However, bladder development and its embryonic function remain poorly understood. Here, we report that in both humans and mice the bladder emerges from the cloaca as the rostral outgrowth, which is flanked by the umbilical arteries. Arrest of the outgrowth, as seen in Shh-null mouse embryos, causes bladder agenesis and unexpected severe defect of the umbilical arteries, suggesting that the outgrowth epithelial signal Shh coordinates development of both the bladder and umbilical arteries. Wnt2 signaling molecule is restricted to the rostral outgrowth mesenchyme; and its expression is dramatically down-regulated in Shh-null mutants. Furthermore, Wnt2-null mutants develop small bladder and single umbilical artery defects. Findings here uncover an unexpected co-development mechanism of the bladder and umbilical arteries orchestrated by the cloacal outgrowth signals. The co-development paradigm offers a conceptual framework to understand evolution and development of the bladder. Significance StatementThe urinary bladder is found in the eutherian mammals but largely absent from other vertebrates. Innovation of the bladder has been viewed as a necessity to manage liquid metabolic wastes on dry land. In this study, we report that the bladder emerges from the rostral outgrowth of the cloaca. We also report an unexpected developmental link between the bladder and umbilical arteries and, demonstrate that their co-development is orchestrated by the outgrowth epithelial signal Shh and mesenchymal signal Wnt2. Findings here highlight a possibility that, in addition to managing metabolic wastes, normal bladder development is required for growth of mammalian embryos by supporting development of the umbilical arteries.

developmental biology↗

bioRxiv2025

NUAK2 is a therapeutically tractable regulator of RNA splicing and tumor progression in neuroendocrine prostate cancer

Prostate cancer remains a leading cause of cancer-related mortality in men, with aggressive, treatment-emergent androgen receptor (AR)-indifferent subtypes, including double-negative prostate cancer (DNPC) and neuroendocrine prostate cancer (NEPC), posing major clinical challenges due to limited therapeutic options. NUAK family kinase 2 (NUAK2), an AMPK-related kinase, has been implicated in tumor growth and metastatic progression; however, its functional significance and therapeutic potential in advanced prostate cancer remain largely unexplored. Here, we identify NUAK2 as a therapeutically actionable kinase dependency in AR-indifferent prostate cancer. Transcriptomic analyses across independent patient cohorts demonstrated progressive upregulation of NUAK2 with disease progression, with the highest expression in NEPC. Immunohistochemical analysis of clinical specimens further confirmed elevated NUAK2 protein expression in NEPC relative to prostate adenocarcinoma. Genetic loss- and gain-of-function studies established NUAK2 as a functional dependency that promotes tumor cell proliferation, clonogenic growth, and tumor growth in vivo. Mechanistically, integrated proteomic analyses revealed that NUAK2 associates with spliceosomal and RNA-processing machinery, while NUAK2 perturbation induced widespread alterations in pre-mRNA splicing programs involving genes linked to mitotic regulation and oncogenic signaling. Pharmacologic studies identified trilaciclib (G1T-28), a clinically approved CDK4/6 inhibitor, as a functionally relevant NUAK2 inhibitor that directly engages NUAK2, suppresses tumor growth, and enhances the efficacy of platinum-based chemotherapy across multiple preclinical models. Collectively, these findings uncover NUAK2 as a previously unrecognized regulator of RNA splicing and therapeutic vulnerability in AR-indifferent prostate cancer and provide a rationale for repurposing G1T-28 and developing NUAK2-directed therapeutic strategies for aggressive, treatment-refractory prostate cancer.

cancer biology↗

bioRxiv2025

Hyperthermic Seizure Susceptibility and Focal Decreases in Parvalbumin-Expressing Cortical Interneurons in a Mouse Model of PCDH19-Clustering Epilepsy

ObjectiveProtocadherin-19 (PCDH19)-clustering epilepsy (PCE) is a severe genetic epilepsy that manifests with early-onset cluster seizures often triggered by fever, intellectual disability, autistic features, and later neuropsychiatric risk. PCDH19 is an X-linked gene critical for brain development. PCE predominantly affects females and rare mosaic males, but not hemizygous males, likely due to cellular mosaicism arising from random X-inactivation and resultant segregation of wild-type and mutant neurons (so-called cellular interference) during development. We generated a novel PCE mouse model and explored neuronal segregation, seizure susceptibility and cortical interneuron distributions. MethodsFemale Pcdh19+/- mice were crossed with X-GFP males to visualize random X-inactivation patterns. Seizure susceptibility was assessed in juvenile mice using hyperthermia and flurothyl exposure. Behavioral testing evaluated cognitive domains. Interneuron distribution in hippocampus and cortex was examined histologically by immunolabeling and crosses with parvalbumin reporter mice. ResultsJuvenile Pcdh19+/- females lacked spontaneous seizures but displayed lower seizure thresholds and more severe seizures during hyperthermia. Seizure susceptibility did not differ from controls after flurothyl exposure. Pcdh19+/- females also exhibited segregation of GFP+ cells in the cortex, hippocampal CA1 region and medial ganglionic eminence, with a marked reduction of parvalbumin-positive interneurons in the CA1 hippocampal region. Although parvalbumin interneuron density was unchanged in the Pcdh19+/- female cortex overall, localized decreases arose in GFP- (Pcdh19 knockout) cortical stripes. InterpretationJuvenile PCE mice exhibit seizure susceptibility to hyperthermia and disrupted the distribution of parvalbumin-expressing interneurons in the hippocampus and cortex. These findings suggest focal parvalbumin interneuron alterations may contribute to PCE pathophysiology.

neuroscience↗

bioRxiv2025

A BCG toxic effector induces mitochondrial DNA compaction to refrain protective immunity

Innate immune memory typically emphasizes nuclear chromatin remodeling that determines transcriptional memory in monocytes and macrophages. Here, we show that compaction of mitochondrial DNA (mtDNA), governed by mitochondrial transcription factor A (TFAM) abundance and mtATP-dependent remodeling, defines the transcriptional and functional state of trained macrophages. While typical pathogen-associated molecular pattern (PAMPs) triggered mtDNA accessibility, Bacillus Calmette-Guerin (BCG) effector PE18 directly interacted with and activated SLC25A5 to reduce mtATP levels, thereby inhibiting TFAM degradation by AFG3L2. Accumulated TFAM restricts mtDNA accessibility, suppressing the transcription of the mtNd1 and thereby reducing mtNd1-mediated mtROS production, which diminishes the trained capacity of macrophages against secondary infections. Notably, BCG{Delta}pe18 maintains macrophages cytokine transcriptional capacity independent of classical nuclear histone modifications. Moreover, BCG{Delta}pe18 provides robust and lifelong protective immunity against Mycobacterium tuberculosis (Mtb) infection. These findings reveal mitochondrial genome as a regulatory entity in trained immunity, and suggest a potential strategy for improving vaccine performance.

immunology↗

bioRxiv2025

Overcoming Extrapolation Challenges of Deep Learning by Incorporating Physics in Protein Sequence-Function Modeling

Understanding protein sequence-to-function relationship is crucial to assist studies of genetic diseases, protein evolution, and protein engineering. The sequence-to-function relationship of proteins is inherently complex due to multi-site high-dimensional correlation and structural dynamics. Deep learning algorithms such as (graph) convolutional neural networks and recently transformers have become very popular for learning the protein sequence-to-function mapping from deep mutational scanning data and available structures. However, it remains very challenging for these models to achieve accurate extrapolation when predicting functional effect of variants with positions or mutation types not seen in the training data. We propose that incorporating the physics of protein interactions and dynamics can be an effective approach to overcome the extrapolation limitations. Specifically, we demonstrate that physics-based modeling can be used to quantify the energetic effects of mutations and that incorporating these physical energetics directly within the convolution and graph convolution neural networks can significantly improve the performance of positional and mutational extrapolation compared to models without biophysics-inspired features. Our results support the effectiveness of leveraging physical knowledge in overcoming the limitation of data scarcity.

bioinformatics↗

bioRxiv2025

A multi-tissue atlas of genetic regulatory effects in sheep

Characterizing the impact of genomic variants on genome function and ultimately complex traits in livestock is essential for the development of sustainable precision agriculture and comparative genomics. Here, as part of the Farm animal Genotype-Tissue Expression (FarmGTEx) project, we present the pilot phase of the SheepGTEx resource through analyzing 6,761 RNA-sequencing samples of 51 primary tissues in a multi-breed population of sheep. We identify millions of regulatory variants associated with seven types of molecular phenotypes, and fine-map 322,468 primary and 114,017 non-primary effects, revealing a high degree of regulatory allelic heterogeneity. We systematically characterize the pleiotropic effects of these variants on molecular phenotypes, assess their context-specific regulatory patterns across tissues, breeds, sexes, and developmental stages, as well as explore their evolutionary constraints across mammals. Finally, we demonstrate the substantial potential of the SheepGTEx resource (https://sheepgtex.farmgtex.org), by providing examples of regulatory mechanisms underpinning 34 complex traits, population divergence between European and Asian breeds, and adaptive evolution in sheep over the past ten millennia.

genomics↗