Biology subjects

Song, J.

Publications and source records attributed to Song, J..

At least 55 records · Page 3Linked to original sources

bioRxiv2026

Non-invasive MRI mapping of tissue-CSF water exchange reveals glymphatic fluid movement in live human cortex

Efficient metabolic waste clearance, via the postulated glymphatic system, is essential for neural homeostasis. However, direct visualization of tissue-cerebrospinal fluid (CSF) exchange remains limited, leading to ongoing debate in the neuroscientific field. The present work revealed evidence of tissue-CSF water exchange in the live human cortex, by employing a novel MRI technique demonstrating the flux of water molecules across the perivascular interface. We observed robust water exchange inside the cortical ribbon, which was more prominent than white matter and deep brain tissue. We validated that the signal originates from CSF and is independent of cerebral perfusion. Water exchange between tissue and CSF declined with age. Furthermore, we demonstrated for the first time that tissue-CSF exchange was impaired in Alzheimers disease (AD), in particular in regions where the perivascular space is clogged by anti-amyloid immunotherapy.

bioengineering↗

bioRxiv2026

EvoRMD: Integrating Biological Context and Evolutionary RNA Language Models for Interpretable Prediction of RNA Modifications

RNA modifications are essential regulators of post-transcriptional gene expression, influencing RNA stability, localization, translation, and degradation. Determining the specific modification at a given nucleotide is therefore critical for understanding its regulatory role. Most computational approaches treat each modification type as an independent binary task. This strategy provides a macro-level statistical perspective, but it does not reflect that, under a defined biochemical or cellular condition, only one modification type can occur at a specific site. Current mapping assays also report a single observed modification per site, leaving all other types unlabeled rather than truly negative. These properties motivate a framework that can reason over competing modification types. We introduce EvoRMD, a unified model for biologically contextualized and interpretable prediction of RNA modification types. EvoRMD integrates contextual sequence embeddings from a large-scale RNA language model with structured biological metadata--including species, organ, cell type, and subcellular localization. A lightweight attention mechanism highlights informative sequence positions. A shared multi-class classifier then generates a context-conditioned plausibility distribution over eleven modification types (Am, Cm, Um, Gm, D, pseudouridine, m1A, m5C, m5U, m6A, m7G), consistent with the single-positive, multiple-unlabeled nature of existing datasets. Although trained in a multi-class setting, EvoRMD also produces calibrated multi-label predictions through sigmoid-transformed logits, enabling direct comparison with existing single-modification and multi-label methods. EvoRMD achieves strong predictive performance and offers interpretable insights through attention profiles and motif analyses. Together, these components establish a biologically grounded framework for identifying and prioritizing RNA modification types from sequence and context.

bioinformatics↗

bioRxiv2026

Single-cell atlas of pig-to-monkey kidney xenotransplantation reveals macrophage chimerism and an IFN-ε orchestrated graft protective immune niche

The clinical translation of xenotransplantation is constrained by an incomplete understanding of the cellular circuitry governing graft rejection versus adaptation under current immunosuppressive regimens. Here, we present a high-resolution single-cell atlas of pig-to-monkey kidney xenotransplantation in a preclinical model, capturing the early immune dynamics preceding graft loss. Our analysis reveals a landscape dominated by innate immunity, characterized by species-specific distribution of macrophages. We identify recipient-derived macrophage subsets (ACKR1+, FN1+, MARCO+, IDO1+) enriched for multiple immune checkpoint molecules, alongside donor-resident subpopulations (APOE+, CCL2+, IL-1A+) exhibiting pro-inflammatory pathogenic signatures, and leverage their transcriptional profiles to computationally predict candidate therapies (belatacept, abatacept, pexidartinib). Further analysis demonstrates that despite divergent differentiation trajectories, both recipient-derived and donor-resident macrophages converge toward hybrid M1/M2 states at their terminal stages. Notably, we uncover a graft-protective circuit orchestrated by epithelial-derived interferon-epsilon (IFN-{varepsilon}), which specifically engages multiple subsets including IDO1+ macrophages to establish a localized immune-tolerant niche. This study establishes the xenograft epithelium as an active participant in immune modulation via the IFN-{varepsilon} axis and reveals macrophage functional chimerism as a key actionable feature of cross-species immune responses.

genomics↗

bioRxiv2026

Salmonella Typhi asparaginase-dependent activation of GCN2 promotes bacterial killing in murine macrophages

Many intracellular pathogens stimulate host cell stress by directly or indirectly causing an imbalance in host nutrients; depletion of amino acid pools in particular can act as a danger signal to infected cells. Using a restrictive host model of Salmonella enterica serovar Typhi (S. Typhi) infection, we identify early induction the integrated stress response (ISR) by viable bacteria, but not heat-killed bacteria. Genetic deletion of the amino acid sensing ISR kinase GCN2 (also known as EIF2AK4) prevented early ISR activation during S. Typhi infection, and murine macrophages lacking GCN2 show impaired bacterial clearance and decreased cytokine output. Supplementation of wildtype C57BL/6 murine macrophages with only the non-essential amino acid asparagine was sufficient to suppress S. Typhi-induced ISR activation and deletion of S. Typhi ansB, encoding an asparaginase, prevented ISR activation during infection. Pharmacological inhibition of mammalian target of rapamycin (mTOR), the other major amino acid sensing pathway in eukaryotic cells, prevented GCN2 activation and ISR induction in murine macrophages, indicating an upstream role for mTOR in signaling to GCN2. These findings suggest a role for the ISR in macrophage innate immune responses to S. Typhi infection and highlight a potential difference in nutrient-dependent signaling between the S. Typhi-susceptible human host and the restrictive murine host centered around asparagine, mTOR, and GCN2.

immunology↗

bioRxiv2026

ACENet: a graph neural network for predicting enzyme pHmin using structure and sequence insights

The acid activity of enzymes, characterized by the minimum pH at which enzymes remain active (pHmin), is crucial for industrial applications in acidic environments. However, the rational design of acid-active enzymes remains challenging due to limited understanding of sequence-structure- activity relationships under acidic conditions. Here, we propose ACENet, a graph neural network that predicts enzyme pHmin by integrating surface features of protein structures with evolutionary representations derived from the large-scale protein language model ESM-2. ACENet achieved a Pearson correlation coefficient of 0.85 on the test dataset, significantly outperforming other deep learning baseline models and maintains stable pHmin predictions under various conditions. Even on a subset of the dataset with less than 20% homology, the PCC remains above 0.5, with an RMSE (Root mean square error) less than 1.4. ACENet also present excellent performance in the annotation of pHmin for homologous proteins and the predictive screening of minimal active pH in protein mutants. Remarkably, ACENet could identify the catalytic region as key determinants of acid activity through residue-level interpretability analysis. Overall, ACENet accelerates the development of highly efficient biocatalysts for diverse applications where acidic conditions predominate.

bioinformatics↗

bioRxiv2026

Modeling Multi-Modal Brain Connectomes for Brain Disorder Diagnosis via Graph Diffusion Optimal Transport Network

Network analysis of human brain connectivity provides a fundamental framework for identifying the neurobiological mechanisms that cause cognitive variations and neurological disorders. However, existing diagnostic models often treat structural connectivity (SC) as a fixed or optimal topological scaffold for functional connectivity (FC). This consequently overlooks the higher-order dependencies between brain regions that are critical for characterizing pathological alterations. Moreover, the distinct spatial organizations of SC and FC complicate their direct integration, as naive alignment methods may distort the inherent nonlinear patterns of brain connectivity. To address these limitations, we propose the Graph Diffusion Optimal Transport Network (GDOT-Net), which models disease-related topological evolution and achieves precise alignment between SC and FC. Unlike existing diffusion studies, the proposed model introduces an evolvable brain connectome modeling approach to infer the complex topological structure of brain networks, unveiling higher-order connectivity patterns linked to specific neuropsychiatric disorders. Furthermore, GDOT-Net incorporates a Pattern-Specific Alignment mechanism, leveraging optimal transport to align structural and functional topological representations in a geometry-aware manner. To capture nonlinear topological relationships between brain regions, a Neural Graph Aggregator Module was developed, which adaptively learns complex node interaction patterns in brain networks. By leveraging this module, GDOT-Net generates highly discriminative representations that form a robust basis for the precision diagnosis of brain disorders. Experiments on REST-meta-MDD and ADNI demonstrate that GDOT-Net surpasses SOTA methods in uncovering structural-functional misalignments and disorder-specific subnetworks. The source code is publicly available at this Link.

neuroscience↗

bioRxiv2026

Myristoylation licenses disordered viral VP4 protein to anchor to and perforate the membrane through phase separation

The VP4 protein of enteroviruses, such as Coxsackievirus B3, is a small, intrinsically disordered protein essential for perforating the host cell membrane during viral entry. A key feature of VP4 is its N-terminal myristoylation, which is required for infectivity in some enteroviruses but dispensable in others, suggesting a complex and context-dependent role that is not fully understood. The precise biophysical mechanisms by which this lipid anchor enables a disordered protein to breach a membrane remain unresolved. Here, using Coxsackievirus B3 VP4 as a model system and integrating multi-scale molecular dynamics simulations with confocal microscopy, we demonstrate that myristoylation is not a simple membrane tether but a multi-functional regulator that orchestrates VP4 activity through distinct, hierarchical roles. First, it provides the necessary hydrophobic anchor to recruit the disordered VP4 to the membrane interface. Second, the myristoyl group acts as a key molecular driver that promotes the liquid-liquid phase separation of VP4, leading to the formation of dynamic condensates on the membrane surface. These condensates actively remodel the membrane, generating substantial curvature that, in turn, lowers the free energy barrier for VP4 penetration. Furthermore, we find evidence that the myristoyl group plays a third role in stabilizing the final transmembrane pore. Our findings establish a novel paradigm where a single lipid modification empowers a disordered viral protein to form a functionally potent condensate that mechanically primes and physically breaches the target membrane, a mechanism that may explain the conditional myristoylation requirement across enteroviruses.

biophysics↗

bioRxiv2026

MolX: A Geometric Foundation Model for Protein-Ligand Modelling

Understanding how small molecules interact with protein binding pockets is central to structure-based drug discovery. Accurately modelling these interactions requires capturing the 3D geometry and physicochemical complementarity of binding interfaces, yet existing computational approaches encode proteins and ligands separately or rely on simplified structural representations that do not explicitly model cross-entity spatial relationships. Such decoupled representations restrict their capacity to capture interface-level geometric constraints that arise from protein-ligand coorganisation. Here we present MolX, a Graph Transformer foundation model that jointly learns geometric and chemical representations of protein pockets and ligands from large-scale 3D structural data. Integrating over 3 million protein pockets and 5 million molecules, MolX represents both entities as E(3)-equivariant graphs to preserve spatial geometry and chemical context. The architecture employs dual E(3)-equivariant graph Transformer encoders to model pocket and ligand embeddings, ensuring representations remain invariant to rotation, translation, and reflection. MolX is pretrained using a hybrid learning paradigm that combines supervised biochemical objectives, logP and energy-gap regression, with self-supervised geometric objectives, coordinate re-construction, and atom-type prediction, fostering generalisable molecular understanding. Across eight downstream benchmarks, including antibody-drug conjugates (ADC), proteolysis-targeting chimeras (PROTAC), molecular glue, and PCBA activity prediction, as well as binding affinity and physicochemical property regression, MolX achieves consistent state-of-the-art performance and strong cross-domain generalisation. Furthermore, MolX incorporates a sparse autoencoder module to decompose latent representations into interpretable biological components, thereby revealing the pocket-ligand interactions that drive prediction outcomes. Together, MolX establishes a scalable and interpretable foundation model for molecular representation learning, providing a unified framework for predicting and interpreting complex small-molecule-protein interactions.

bioinformatics↗

bioRxiv2026

NRF2 pathway activation and SPP1⁺TREM2⁺ macrophages drive chemoradiotherapy resistance in esophageal squamous cell carcinoma

Esophageal squamous cell carcinoma (ESCC) is among the most aggressive cancers, with low rates of durable response to chemoradiotherapy and limited therapeutic options for relapsed disease. To uncover mechanisms of treatment resistance and relapse, we performed comprehensive multi-omics profiling of >100 pre-treatment and post-relapse ESCC tumors from a prospective clinical trial (NCT04694391), integrating whole-exome/genome sequencing, bulk RNA-sequencing, single-cell RNA sequencing, and spatial transcriptomics. We identify somatic alterations in NFE2L2/KEAP1 in nearly 40% of relapsed patients, which are associated with upregulation of NRF2 signaling targets in resistant tumors and cell line models. At single-cell and spatial resolution, relapsed tumors are enriched for NRF2-activated epithelial cells that physically co-localize with immunosuppressive SPP1TREM2 macrophages. This co-localization suggests a synergistic interaction between NRF2-driven tumor programs and macrophage-mediated immune suppression that promotes relapse after chemoradiotherapy. Our findings nominate NFE2L2/KEAP1 mutations as predictive biomarkers for patient stratification and highlight therapeutic targeting of NRF2 signaling and SPP1TREM2 macrophages as rational strategies to overcome resistance in ESCC.

cancer biology↗

bioRxiv2026

OT-knn: a neighborhood-aware optimal transport framework for aligning spatial transcriptomics data

Spatial transcriptomics (ST) measures gene expression while preserving spatial context within tissues, enabling detailed characterization of tissue organization. As ST technologies advance, aligning datasets across tissue sections, individuals, platforms, and developmental stages has become increasingly important but remains challenging due to sparse expression, biological heterogeneity, and geometric distortions between slices. We introduce OT-knn, a method for ST alignment that integrates local neighborhood information within an optimal transport framework. Rather than relying solely on single-spot expression, OT-knn reconstructs each spot using its spatial k-nearest neighbors, capturing microenvironment context that is more robust to noise and variability. These representations are then used to derive probabilistic correspondences between slices. We evaluate OT-knn using simulated data with known ground-truth alignment and real datasets from multiple ST platforms, including human dorsolateral prefrontal cortex data (10x Genomics Visium), mouse brain aging data with both within-donor and cross-donor comparisons (MERFISH), a multi-stage axolotl brain dataset (Stereo-seq), and a cross-platform analysis using mouse embryo datasets profiled by Stereo-seq and seqFISH. Across these settings, OT-knn achieves accurate and robust alignment, particularly in the presence of spatial deformation, donor heterogeneity, developmental variation and technological differences.

bioinformatics↗

bioRxiv2026

ProteinConformers: large-scale and energetically profiled descriptions of protein conformational landscapes

Modeling protein conformational landscapes is essential for understanding dynamics, allostery, and drug discovery, yet existing resources lack diverse conformational coverage, energetic annotations, or benchmarking standards. ProteinConformers (https://zhanggroup.org/ProteinConformers) provides 2.7 million geometry-optimized conformations generated with a multi-seed molecular dynamics strategy, paired with 13.7 million energy evaluations and 5.5 million similarity annotations. It delivers continuous landscapes from non-native to near-native states, benchmarking framework for multi-conformation generators, and an interactive analysis platform.

bioinformatics↗

bioRxiv2026

Heterogeneous and specific localization of mRNAs at dendritic spine and its dependence on microtubule entry and kinesin motor

Local mRNA translation near neuronal synapses is crucial for structural plasticity during brain development and memory formation. Whereas mRNAs are transported from soma to dendrites as individual ribonucleoprotein (RNP) granules, how do they subsequently localize at synapses is unclear. In particular, whether different dendritic mRNAs might display differential synaptic localization and interact with specific postsynaptic proteins remains unexplored. Here we found that two well-known dendritic mRNAs, Actb and Camk2a, are targeted to different subdomains of dendritic spines and delivered to different spine populations after synaptic stimulation. Surprisingly, the induced mRNA localization to dendritic spines is associated with microtubule (MT) entry and specifically requires the MT-associated motor KIF5A. Proximity-labelling proteomics further identifies distinct postsynaptic proteins enriched by Actb and Camk2a, from which we uncover the cytoskeletal protein Septin7 to be co-localized with the two mRNAs at different subdomains of dendritic spines. Our study reveals an unexpectedly specific mRNA localization mechanism at synapses and demonstrates proximity-labelling as a useful approach to map their synaptic partners.

neuroscience↗

bioRxiv2026

Interspecies interactions drive community-level selection in microbial coalescence

It has long been debated in ecology whether communities behave as cohesive units or as loose collections of independent species. Here, we study this question in the context of community coalescence, the mixing of previously isolated communities, using bacterial microcosm experiments combined with ecological modeling. Our results demonstrate that interspecies interaction strength determines whether communities or species are the units of selection during coalescence. When interactions are moderate to strong, one parental community consistently outcompetes the other, indicating community-level selection. In contrast, under weak interactions, species fates are uncorrelated and the two communities contribute equally to the coalesced outcome, indicating the absence of community-level selection. These patterns extend to communities derived from natural samples with greater taxonomic diversity and richness. Furthermore, we identify two distinct regimes underlying community-level selection in experiments with different media conditions: an emergent regime in which collective dynamics shape outcomes that cannot be predicted from species traits alone, and a top-down regime where dominant species determine the winning community. Together, these results reconcile conflicting observations on community-level selection during community coalescence by demonstrating that communities behave as cohesive units only when interactions are sufficiently strong.

ecology↗

bioRxiv2026

xTracer: Integrating chromatogram and mobiligram correlations for untargeted peptide identification in SLIM-based PAMAF data

Parallel Accumulation with Mobility-Aligned Fragmentation (PAMAF) achieves near-complete ion utilization and high spectral specificity by fragmenting all mobility-separated precursors without quadrupole isolation. Leveraging the ultrahigh mobility resolution of SLIM, this quadrupole-free strategy maximizes ion utilization efficiency and offers a promising approach in mass spectrometry-based proteomics, particularly for low-abundance peptides or low-input samples. However, the unique data structure of PAMAF where precursor-fragment relationships are encoded along the mobility dimension renders it incompatible with existing peptide identification tools. Here, we present xTracer, the first untargeted peptide identification algorithm developed specifically for PAMAF data. xTracer integrates correlations across both chromatographic and mobility dimensions to associate precursor and fragment ions, reconstruct pseudo-spectra, and enable database searching using well-established DDA search engines. Applied to datasets with varying sample loads and acquisition throughputs, xTracer consistently achieved robust and reproducible peptide identifications, outperforming single-domain correlation strategies. Overall, xTracer provides a versatile and high-efficiency computational framework for reconstructing pseudo-spectra from quadrupole-free, mobility-aligned fragmentation data, enhancing the analytical power of high-resolution ion mobility (HRIM)-based proteomics.

systems biology↗

bioRxiv2026

Correlations of 276 missense variants of PSEN1, PSEN2, and APP on the production of Aβ peptides against variant effect predictors and biophysical structures

INTRODUCTIONEarly-onset Alzheimers disease (EOAD) is mostly caused by mutations in Presenilin-1,2 (PSEN1, PSEN2), and amyloid precursor protein (APP) genes. Both genetic and experimental evidence has supported the PSEN-APP amyloid hypothesis as the leading pathogenesis of AD. Thus far, 276 missense variants from PSEN1,2, and APP have been identified. Herein, we report the development of algorithm platforms for predicting the degrees of pathogenic potentials of these various missense mutations in AD. METHODSWe performed extensive pairwise correlations between 276 in vitro functional assays of missense variants from PSEN1, PSEN2, and APP and clinical data among EOAD patients against 37 variant effect predictors (VEPs). Furthermore, we used the structural models of all three proteins to refine understanding of the molecular basis of missense variants in dominant negative (DN) or loss-of-function (LoF) effects. RESULTSWe found that VEPs were consistently correlated with age of onset (AAO) among EOAD patients and the A{beta}42/A{beta}40 ratio biomarker. This study also identified discrepancies in the predictor variable leading to an increase in the A{beta}42/A{beta}40 ratio; The increased A{beta}42/A{beta}40 ratio was due to decreased A{beta}40 levels by Sun et al., (2016), while Petit et al., (2022) argued that increased A{beta}42 level was the culprit. We also identified that with increased predicted pathogenicity, there were decreased A{beta}38 levels. Conversely, this study found no direct correlations with A{beta}37 and A{beta}43. Lastly, structural studies show characteristics of either DN and LoF mechanisms, as there were numerous variants located in buried hydrophobic domains and those on charged surfaces of the proteins. DISCUSSIONCurrently, a major limitation in the field is the lack of reliable approaches for predicting the pathogenicity of variants in EOAD. We show that VEPs are promising in deciphering the effects of variants in PSEN1, PSEN2, and APP genes, especially between multiple pathogenic mutations. Our study also illuminates the need for experimental studies to identify the predictor variable causing the increased A{beta}42/A{beta}40 biomarker. Our biophysical studies also identify the need to characterize whether the mechanism of pathogenesis caused by DN or LoF effects is in scope of the presenilin hypothesis. Research in ContextO_LISystematic review: The authors reviewed experimental evidence of pathogenic PSEN1, PSEN2, and APP variants in scope of the presenilin hypothesis. To our knowledge, the pathogenesis of EOAD remains incompletely understood. There are also no predictive tools/models available for informing clinical translation. C_LIO_LIInterpretation: VEPs demonstrate strong correlations with AAO and A{beta} levels from pairwise correlations with 276 missense variants. Structurally, it appears that mutations demonstrate characteristics of dominant negative and loss-of-function effects. C_LIO_LIFuture Direction: Larger studies are necessary of VEPs for realistic clinical usage. C_LI HighlightsO_LIWe performed correlations of 37 VEPs against in vitro data of EOAD variants. C_LIO_LIVEPs demonstrate potential as diagnostic tools for early onset Alzheimers disease. C_LIO_LIPathogenic missense variants show characteristics of DN and LoF effects. C_LI

neuroscience↗

bioRxiv2026

Integrative Multi-Omics Analysis Identifies a Functional Enhancer Driving Tumorigenesis in Head and Neck Squamous Cell Carcinoma

Enhancer is a critical epigenetic feature in head and neck squamous cell carcinoma (HNSCC), yet the functional roles of individual enhancers remain poorly understood. Here, we conducted an integrative multi-omics analysis based on publicly available ATAC-seq, H3K27ac ChIP-seq, transcriptomic profiling, and genetic association datasets to systematically map HNSCC-associated enhancers and their target genes. Integrating ATAC-seq and H3K27ac ChIP-seq, we identified 20,362 enhancers and 18,040 enhancer-associated genes, highlighting widespread regulatory complexity. Functional characterization of the TERT-associated enhancer GH05J001312 (GRCh38/hg38: chr5:1312099-1317743) revealed strong transcriptional activity in HNSCC. CRISPR-mediated deletion of a core sequence significantly reduced TERT expression, impaired cellular proliferation in vitro, and suppressed tumor growth in vivo, confirming its role as a key cis-regulatory element. RNA-seq analysis of enhancer-edited cells uncovered 742 differentially expressed genes enriched in cancer-related pathways, including MAPK and IL-17 signaling, indicating a broad transcriptional impact. Collectively, our findings establish GH05J001312 as a functional enhancer driving oncogenic programs in HNSCC and suggest enhancer-targeted strategies as a potential therapeutic avenue.

cancer biology↗

bioRxiv2026

PRISM: Niche-informed Deciphering of Incomplete Spatial Multi-Omics Data

Spatial multi-omics data, characterizing the knowledge of diverse molecular layers, have become indispensable for the in situ analysis of tissue architecture and complex biological processes. Nevertheless, current spatial multi-omics sequencing protocols are often hindered by incompatible protocols, resulting in incomplete spatial multi-omics pairing due to inconsistent field-of-view or spatial resolution. To address this, we present PRISM, a novel computational method tailored for this scenario. PRISM leverages a niche-informed graph to propagate information from paired to unpaired regions, jointly achieving the spatial domain identification and spatial omics imputation. Extensive benchmarking on five diverse simulated and real experimental datasets demonstrated that PRISM outperformed existing methods in spatial multi-omics analysis tasks. Application to the human Parkinsons disease data revealed that PRISM accurately recovered dopamine-associated spatial domains and metabolite distributions masked by incomplete data gaps. PRISM offers a robust solution for bridging the integration gap inherent in incompatible sequencing protocols, thereby facilitating more accurate downstream biological interpretation.

bioinformatics↗

bioRxiv2026

De novo design of a safe and potent respiratory syncytial virus immuno-focusing antigen

Respiratory syncytial virus (RSV) remains the leading cause of severe respiratory infections in infants, the elderly, and the immunocompromised. Although stabilized full-length pre-fusion (pre-F) protein vaccines are promising, enhanced respiratory disease (ERD) remains a critical safety concern. Here, we used artificial intelligence to design a de novo immuno-focused antigen that structurally preserves the RSV F head region containing critical neutralising epitopes-- site O, II and V-while replacing the non-neutralising stem with a computationally designed scaffold to minimise immunopathological risk. The lead candidate, aRF6, elicited robust protective immunity against RSV in mice and similar immunogenicity in non-human primates without detectable toxicity. Importantly, in stringent ERD-promoting models, aRF6 induced minimal pulmonary pathology and markedly attenuated Th2-biased cytokine responses, outperforming formalin-inactivated RSV and full-length-stabilized pre-F. The results of cryoelectron microscopy confirmed that the aRF6 structure precisely matched the computational predictions. These results demonstrated that computationally designed de novo immuno-focused antigens can yield safe and effective RSV vaccines, thereby providing a rational framework for next-generation vaccine development.

microbiology↗