Biology subjects

Peng, Z.

Publications and source records attributed to Peng, Z..

At least 37 records · Page 2Linked to original sources

bioRxiv2026

Lung-Selective Immune Reprogramming via In Situ Red Blood Cell Hitchhiking Nanoparticles

Premature clearance and limited organ targeting remain major barriers for nanoparticle (NP) drug delivery. Hitchhiking NPs on red blood cells (RBCs) can enhance circulation and organ-selective accumulation, but most approaches require ex vivo RBC extraction and reinfusion, limiting clinical translation. Here, we report an in situ RBC-hitchhiking strategy, named i-Bind, which employs polyphenol surface functionalization to enable spontaneous NP attachment to RBCs directly in the bloodstream. Driven by strong interactions of phenolic motifs with RBC membranes, i-Bind NPs exhibited markedly enhanced and more stable hitchhiking onto RBCs under flowing whole blood conditions. In both healthy and diseased mice, i-Bind NPs preferentially target the lungs, resulting in an over 20-fold increase in lung-to-liver deposition ratio compared to unmodified NPs. Additionally, i-Bind NPs show preferential targeting to distinct lung immune cell subsets in a pathology-dependent manner, including cDC2s in healthy lungs, neutrophils in acute lung injury, and cDC1s in lung metastases. In a melanoma lung metastasis model, delivery of the STING agonist diABZI via i-Bind NPs significantly inhibited lung metastasis progression by reprogramming the lung immune microenvironment. Collectively, i-Bind provides a simple and versatile platform for organ-targeted drug delivery and immune reprogramming. TeaserSurface functionalization of nanoparticles enables in situ red blood cell hitchhiking, unlocking new paths for organ-selective immune reprogramming

bioengineering↗

bioRxiv2026

T2T Genome and Population Resequencing Reveal OfCCD4 Alleles Orchestrate Petal Color and Scent in Osmanthus fragrans

Osmanthus fragrans is prized for its floral color and fragrance, both key targets for genetic improvement. However, the lack of a complete genome assembly and comprehensive population structure hinders gene dissection and marker development for breeding. To investigate the genetic basis of color-scent interaction, a telomere-to-telomere (T2T) genome assembly is generated, and whole-genome resequencing of 100 cultivars is performed. Integrative population and metabolomic analyses reveal a trade-off: orange-red ( Aurantiacus) cultivars accumulate high /{beta}-carotene but low aromatic apocarotenoids (/{beta}-ionone), while yellow-white cultivars show the opposite pattern. Divergence mapping identifies OfCCD4 as the major underlying locus. Three alleles are characterized--functional (A), partially functional (aDel), and a frameshift null (aStop)--with aStop strictly co-segregating with the Aurantiacus phenotype. Transient and stable transformations confirm that A and aDel cleave /{beta}-carotene into /{beta}-ionone, while aStop abolishes this activity. A co-dominant PCR marker is developed based on allele-specific polymorphisms. OfCCD4 is thus established as the key regulator of the color-scent trade-off in osmanthus. The identified alleles and molecular marker enable rapid, low-cost genotyping at the seedling stage--offering particular value for marker-assisted selection in ornamental breeding, where extended juvenility makes phenotypic selection highly time- and resource-intensive.

plant biology↗

bioRxiv2026

Genomic insights into rapid diversification and adaptive introgression in the earthworm Amynthas aspergillum

How species diversify in highly fragmented landscapes while maintaining genetic connectivity remains a central question in evolutionary biology. The megascolecid earthworm Amynthas aspergillum exemplifies this paradox, showing deeply divergent mitochondrial lineages despite extensive nuclear gene flow across southern China. We conducted whole-genome resequencing across its core distribution in the Guangxi karst region to elucidate the population structure and processes underlying this cytonuclear discordance. Genome-wide analyses identified three genetically distinct groups shaped by fine-scale karst isolation, alongside pervasive historical and contemporary introgression. Demographic modeling revealed rapid diversification during the Late Pleistocene ([~]0.14LJMa) with pronounced fluctuations in effective population size. Although isolation increased inbreeding and mutational load in some lineages, introgression persisted even among deeply divergent mitochondrial clades. Notably, several introgressed regions showed strong signatures of positive selection, including genes involved in mitochondrial function and metabolism.Together, these results demonstrate that gene flow can simultaneously mitigate the genomic costs of isolation and promote functional innovation via adaptive introgression, providing a genomic framework for diversification with gene flow in subterranean invertebrates inhabiting fragmented karst landscapes.

genomics↗

bioRxiv2026

PathDiffusion: modeling protein folding pathway using evolution-guided diffusion

Despite remarkable advances in protein structure prediction, a fundamental question remains unresolved: how do proteins fold from unfolded conformations into their native states? Here, we introduce PathDiffusion, a novel generative framework that simulates protein folding pathways using evolution-guided diffusion models. PathDiffusion first extracts structure-aware evolutionary information from 52 million predicted structures the AlphaFold database. Then an evolution-guided diffusion model with a dual-score fusion strategy is trained to generate high-fidelity folding pathways. Unlike existing deep learning methods, which primarily sample equilibrium ensembles, PathDiffusion explicitly models the temporal evolution of folding. On a benchmark of 52 proteins with experimentally validated folding pathways, PathDiffusion accurately reconstructs the order of folding events. We further demonstrate its versatility across four challenging applications: (1) recapitulating Antons molecular dynamics trajectory for 12 fast-folding proteins, (2) reproducing functionally important local folding-unfolding transitions in 20 proteins, (3) characterizing conformational ensembles of 50 intrinsically disordered proteins, and (4) resolving distinct folding mechanisms among 3 TIM-barrel proteins. We anticipate that PathDiffusion will be a valuable tool for probing protein folding mechanisms and dynamics at scale.

bioinformatics↗

bioRxiv2025

Gut microbial interaction networks control autoimmunity to neuroretina

The gut microbiome influences the development of immune-mediated inflammatory diseases, including autoimmune uveitis, a sight-threatening ocular inflammation driven by retina-specific T cells1. Using a model of spontaneous autoimmune uveitis (sEAU) we showed that gut commensals provide immune stimuli that trigger disease2. Here we report that uveitis-promoting microbes are present in human gut flora and that colonization of germ-free (GF) mice with commensals from healthy human donors was sufficient to provoke disease. Severity of sEAU correlated with expansion of Akkermansia and contraction of short-chain fatty acid (SCFA)-producing Firmicutes, followed by decreased SCFA levels and a dominant gut Th1 effector response. Mechanistic gain-of-function experiments, enriching GF sEAU mice with Akkermansia, reproduced these microbiome, metabolite and immune phenotype shifts, and exacerbated disease, suggesting that Akkermansia promotes autoimmunity by outcompeting SCFA-producers and enhancing Th1-type responses. An inverse correlation between Akkermansia (Verrucomicrobia) and Firmicutes was also present in patients with uveitis, multiple sclerosis and Crohns disease. These findings reveal a stereotypic gut microbial interaction network that regulates systemic immune balance, and may represent an ecologically conserved mechanism through which the gut microbiome modulates autoimmune and inflammatory diseases.

immunology↗

bioRxiv2025

Siglec-engaging immunosuppressive sialoglycans are upregulated in prostate cancer and are targetable to suppress bone metastasis

Prostate cancer is a leading cause of male cancer-related deaths over the age of 50. New treatment options for prostate cancer are urgently needed, especially for tumours that have spread to bone. Aberrant sialylation holds substantial potential for the discovery of new therapeutic targets but has remained relatively unexplored in the context of prostate cancer, primarily due to the lack of reliable reagents for detecting tumour sialoglycans in clinical tissue. Here, we address this knowledge gap using high-affinity Siglec-based sialoglycan-binding reagents (HYDRAs) to quantify tumour sialoglycans in tissues representing the full clinical heterogeneity of prostate tumours. Using HYDRA immunohistochemistry, we show that sialoglycans that can engage Siglec-3, -7, and -9 are upregulated in primary prostate cancer tissue and sialoglycan ligands for Siglec-7 correlate with prostate cancer bone metastasis and poorer patient prognosis. Analysis of prostate-derived tumours growing in bone reveals Siglec receptors are expressed by immune cells in the bone metastatic tumour microenvironment, suggesting that this axis may play a role in immune cell functions in bone metastatic prostate cancer. Indicating this is clinically actionable, an engineered bisialidase (E-612) can effectively strip Siglec ligands from prostate cancer cells and prolong survival times of mice with bone metastasis. Our findings identify a novel mechanism involving Siglec-engaging sialoglycans in driving the growth of prostate cancer bone metastasis and demonstrate how this axis can be targeted to impede lethal prostate cancer progression.

cancer biology↗

bioRxiv2025

Aberrant 5-Methylcytosine tRNA Modification Disrupts Proteostasis and Exacerbates Age-Related Osteoporosis

Age-related protein aggregation drives senile osteoporosis. Aberrant tRNA modifications exacerbate the progression, yet mechanisms linking these to bone loss remain unclear. In this study, we identify Nsun2 (m5C methyltransferase) as key regulator: age-dependent Nsun2 downregulation in BMSCs reduces mC levels of tRNAs, destabilizing tRNAs and impairing translation efficiency of specific transcripts. This directly disrupts the protein synthesis of molecular chaperone and pro-osteogenic factors, accelerating misfolded protein aggregates and activating unfolded protein response, inducing BMSCs senescence and impairing osteogenesis. Mice specifically depleted of Nsun2 exhibited reduced bone mass, whereas mice overexpressing Nsun2 alleviated age-associated bone loss. Notably, the exacerbated protein aggregation and bone mass loss in Nsun2-deficient mice were ameliorated following treatment with the molecular chaperone activator-ML346. Remarkably, ML346 administration proved sufficient to reverse age-related functional deficits in aged mice. Overall, our findings demonstrate that aberrant tRNA-m5C modification alters protein synthesis and induces proteostasis collapse, which constitute a novel contributor to the pathogenesis of senile osteoporosis. Additionally, reduction of protein aggregation through the activation of molecular chaperones presents a promising therapeutic strategy for this disease.

cell biology↗

bioRxiv2025

Cell cycle-driven epigenetic resetting maintains stem cell fate for shoot branching

Adult mammalian stem cells typically maintain stem cell identity through proliferative quiescence. In contrast, we demonstrate that stem cell maintenance in Arabidopsis bud precursor cells requires active cell-cycle progression. Inhibiting division silences the shoot meristem marker gene SHOOT MERISTEMLESS (STM) and promotes differentiation. Whereas proliferation dilutes H3K27me3 levels to counteract silencing. Meanwhile, we identified two classes of transcription factors recruiting polycomb repressive complex 2 (PRC2) to epigenetically silence STM. The balance between these forces establishes a cell cycle-coupled epigenetic "Sisyphus" mechanism that maintains pluripotency. This cell fate switch is bistable; modeling and experimental data confirm that prolonged quiescence triggers irreversible differentiation. We propose that sequence-dependent PRC2 recruitment in plants enables precise silencing of fate-determining genes, while cell proliferation sustains pluripotency by resetting epigenetic marks.

plant biology↗

bioRxiv2025

Effects of different proportions of ground corn and steam-flaked corn diets on growth performance, serum biochemical parameters, rumen fermentation and rumen microbial metagenome of yaks

This study objected to assess the effect of substituting ground corn (GC) with steam-flaked corn (SFC) at various ratios on growth performance, serum biochemistry, rumen fermentation, and microbiome in fattening yaks. Fifty male Maiwa yaks (196.43 {+/-} 16.08 kg) were randomly assigned to five groups: SFC 0, SFC 25, SFC 50, SFC 75, and SFC 100, representing 0%, 25%, 50%, 75%, and 100% substitution of GC with SFC. Results showed that the dry-matter intake (DMI) in SFC 0 group was significantly lower than those in other treatment groups (P < 0.05). Among the serum biochemical parameters, the triglyceride (TG) content in the SFC 100 group was significantly higher than other treatment groups (P < 0.05), with a quadratically increase (P < 0.05). Regarding rumen fermentation parameters, the propionate in the SFC 50 and SFC 100 groups was significantly higher than in the other groups (P < 0.05). At the phylum level of the rumen microbiome, the replacing with SFC significantly increased the relative abundance of Planctomycetota, Bacteria, and Candidatus_Saccharibacteria (P < 0.05). At the genus level, the replacing with SFC significantly increased the relative abundance of Paludibacteraceae and Bacteria (P < 0.05). Among the CAZy enzymes in the rumen, the relative abundance of cellulase modules (CM) in the SFC 0 group was significantly higher than other groups (P < 0.05). In conclusion, replacing 50% of GC with SFC may be an effective strategy to enhances growth performance in yak farming by improving rumen fermentation, microbiome composition, and carbohydrate-active enzymes. HighlightsO_LIThe replacement of GC with SFC can optimize rumen fermentation of yaks. C_LIO_LIThe replacement of GC with SFC in the diets affected the rumen microbiome of yaks. C_LIO_LIMetagenome profiled differences between CAZy enzymes and eggNOG functional proteins. C_LI

zoology↗

bioRxiv2025

Quality assessment of RNA 3D structure models using deep learning and intermediate 2D maps

Accurate quality assessment is critical for computational prediction and design of RNA three- dimensional (3D) structures. In this work, we introduce RNArank, a novel deep learning-based approach to both local and global quality assessment of predicted RNA 3D structure models. For a given structure model, RNArank extracts a comprehensive set of multi-modal features and processes them with a Y-shaped residual neural network. This network is trained to predict two intermediate 2D maps, including the inter-nucleotide contact map and the distance deviation map. These maps are then used to estimate the local and global accuracy (i.e., lDDT). Extensive benchmark tests indicate that RNArank consistently outperforms traditional methods and other deep learning-based methods. Moreover, RNArank demonstrates promising performance in identifying high-quality structure models for targets from the recent CASP15 and CASP16 experiments. We anticipate that RNArank will serve as a valuable tool for the RNA biology community, improving the reliability of RNA structure modeling and thereby contributing to a deeper understanding of RNA function.

bioinformatics↗

bioRxiv2025

Conditions enabling the persistence of cooperating ribozymes without cellular encapsulation

Explaining the origin of molecular systems composed of cooperating polymer sets that confer both metabolic and information processing functions is a key challenge in origins-of-life research. A related puzzle is the emergence of polymers of sufficient length to confer complex functions such as the RNA-dependent RNA polymerization with proofreading. Addressing these issues using computational models of template-guided replicating polymer systems is generally constrained by the exponential increase in diversity as the length of polymers increases. In this study, inspired by the computer game Tetris(R) and the Polymerase Chain Reaction (PCR) technique, we developed an abstract computational model of cooperative replicating polymer systems that avoids tracking all potential sequences. Using this model, we explored cooperative chemical ecosystems consisting of catalytic polymers conferring functions analogous to kinases, ligases, and mutation inhibitors. We show that prebiotic environments with micro-compartments with local exchanges enable multilevel selection that facilitates the survival of cooperating polymers. The ability of cooperative systems to persist is sensitive to intrinsic properties of catalysts such as catalytic efficiency and extrinsic factors such as dilution rate. These results provide a roadmap for future studies that look not just at persistence but also at the stepwise, de novo emergence of chemical ecosystems with both metabolic and information-processing capabilities.

evolutionary biology↗

bioRxiv2025

Proteomic characterization of major fish allergy responsive protein parvalbumins in Hilsa (Tenualosa ilisha): A commercially important fish in Southeast Asia

The protein parvalbumin (PRV)-beta (PRVB) is the primary cause of food allergies to bony fish. Although PRVB is a well-characterized protein in many bony fishes, little is known about the Hilsa, an anadromous fish with great economic importance and predominantly found in Southeast Asia. In this study, we characterized the Hilsa PRV utilizing various proteomic approaches in response to two major riverine habitats and developmental stages. Unique peptide sets corresponding to three PRV isoforms were identified in Hilsa muscle tissues. Label-free quantitative proteomic analysis coupled with ELISA revealed higher levels of PRVB in young fish compared to adults, irrespective of their riverine habitats. A comparative quantitative analysis of PRVB further demonstrated that Hilsa had less PRVB than other commonly consumed freshwater fish species. Multiple reaction monitoring (MRM)-based targeted proteomic approach showed the potential of PRV as a marker protein for allergen quantitation and authenticating the presence of Hilsa in a complex freshwater fish mixture. Our findings collectively offer fundamental knowledge on Hilsa PRVs for further investigation on the food safety and quality evaluation of Hilsa fish.

biochemistry↗

bioRxiv2025

The Neurofibrillary Tangle Maturity Scale: A Novel Framework for Tangle Pathology Evaluation in Alzheimer's Disease

Neurofibrillary tangles are dynamic neuropathologic hallmarks of Alzheimers disease with a hypothesized lifespan morphologically-defined by three maturity levels: pretangles, mature tangles, and ghost tangles. To better understand the progression of tangle pathophysiology, we characterized tangle maturity level predilection of 15 tau antibodies recognizing a broad range of linear, phosphorylation, conformational, and truncation epitopes in the hippocampus of 24 postmortem brains. We developed the tangle maturity scoring system to semi-quantitatively evaluate each tangle maturity level. Based on proportions of tangle maturity levels, we classified antibodies as "early" (mostly pretangles and mature tangles), "middling" (mature tangles with pretangles and ghost tangles), and "advanced" (mostly ghost tangles and mature tangles) tangle maturity markers. To summarize tangle maturity predilection, we developed the tangle maturity scale to integrate individual tangle maturity scores. Correlations showed stronger relationships between tangle maturity scale and subsector thickness for more advanced tangle maturity markers in CA1 and subiculum, whereas Braak tangle stage remained consistently correlated throughout markers of the tangle lifespan. To aid in scoring hippocampi, we used machine learning to recognize tangle maturity levels, which performed comparably to a domain expert and showed similar relationships by Spearman correlation. Pattern recognition software was used to assess tangle and neuritic tau burden separately, which generally correlated with Braak stage and neuronal counts. However, tangle-derived tau burden more consistently correlated with hippocampal subsector thickness. In conclusion, we developed manual and automated scoring systems to evaluate tangle maturity levels, demonstrating early 4R, phosphorylated, and oligomeric tau accumulation preceding more advanced 3R and truncated tau. Our study provides supportive evidence of disease-relevant ordering of tau posttranslational modifications in the brain, which may have implications for theragnostic development. These findings underscore the promise of computerized quantitative analyses (i.e., pathomics) for high-throughput feature extraction from whole-slide images to enhance our understanding of microscopically observed morphologic changes.

neuroscience↗

bioRxiv2025

Generating dynamic structures through physics-based sampling of trRosettaX2-predicted inter-residue geometries

Deep learning methods like AlphaFold2 have achieved remarkable breakthroughs in predicting static protein structures. However, accurately modeling alternative conformations and dynamic structures remains a significant challenge. Here, we introduce trRosettaX2-Dynamics (trX2-D), a novel approach to addressing this challenge through physics-based iterative sampling of the trRosettaX2-predicted inter-residue geometric distributions. trX2-D was first pre-trained on high-resolution X-ray structures and then fine-tuned on [~]7,000 dynamic NMR structures, enhancing its inherent ability to predict alternative conformations and dynamic structures. Leveraging a transformer-based neural network, it first predicts an initial set of inter-residue geometric constraints, which are then sampled to generate dynamic structures iteratively, without requiring any prior knowledge of native structural states. Comprehensive benchmarks on three datasets, two for alternative conformations and one for dynamic structures, demonstrate that trX2-D shows promise in predicting alternative conformations and capturing structural dynamics. This work illustrates the potential of combining deep learning predictions with physics-based sampling for advancing the prediction of protein dynamic structures.

bioinformatics↗

bioRxiv2025

Hybrid Deep Learning with Protein Language Models and Dual-Path Architecture for Predicting IDP Functions

Intrinsically disordered regions (IDRs) drive essential cellular functions but resist conventional structural-function annotation due to their dynamic conformations. Current computational methods struggle with cross-dataset generalization and functional subtype discrimination. We present IDPFunNet, a hybrid deep learning model combining convolutional neural network, bidirectional LSTM, residual MLP, and the protein language model ProtT5 to predict six IDR functional classes: five binding subtypes and disorded flexible linkers (DFLs). Its dual-path architecture decouples binding IDR prediction and DFL identification using evolutionary embeddings from ProtT5, which outperformed ESM-family models and AlphaFold2 structural features by [≥] 1.3% average AUC and [≥] 12.7% average APS. Validated across six benchmarks including CAID2/3 blind tests, IDPFunNet achieves superior protein-binding prediction (AUC 0.832-0.866) with significant improvements ([≥]1.5%, p-value<0.05) over existing methods, while matching specialized DFL predictors. Multi-task learning enhances protein/lipid/small molecule-binding performance (3.1-35.1% AUC gains) with BiLSTMs proving optimal DFL identification, despite self-attentions potential for nucleic acid-binding (AUC 0.831). This framework overcomes methodological limitations through interpretable, generalizable IDR functional mapping.

bioinformatics↗

bioRxiv2025

Rapid and accurate protein structure database search using inverse folding model and contrastive learning

Protein structure database search has become increasingly challenging due to the growing number of experimental and computational structures. We introduce mTM-align2, a novel two-step approach for rapid and accurate protein structure database search. In the first step, protein structures are first transformed into embeddings using a pre-trained inverse folding model (ESM-IF) and 3D Zernike polynomials. The ESM-IF embeddings are further optimized through a contrastive learning network, which is trained on [~]7 million structure pairs. Structures with similar embeddings are returned on the fly in this step. The second step employs a rapid structure alignment program to refine top candidates, ensuring high precision and producing high-quality alignments. Extensive benchmarks reveal that mTM-align2 performs competitively compared to other leading methods, completing monomeric structure search in seconds with over 90% precision for the top 10 hits. The t-SNE visualization of the mTM-align2 embeddings for thousands of structures demonstrates that our embeddings are structurally informed, capturing the global structural features. It uncovers insights such as structure misclassifications and ambiguous structural class boundaries. A web server for mTM-align2 is accessible at https://yanglab.qd.sdu.edu.cn/mTM-align/.

bioinformatics↗

bioRxiv2025

A Non-canonical Role for Hepatocyte MLKL in Promoting Mitochondrial Dysfunction and Senescence in the Aging Liver

Liver aging is characterized by chronic inflammation and metabolic dysfunction that contributes to the progression of metabolic dysfunction-associated steatotic liver disease (MASLD). Necroptosis, a form of inflammatory cell death, is activated in aging livers, and genetic (Ripk3-/-or Mlkl-/- mice) or pharmacological (RIPK1 inhibitor necrostatin-1s) inhibition of necroptosis attenuates liver inflammation and pathology. However, the cell type-specific role of necroptosis in liver aging remains unclear. Given that MLKL is expressed in hepatocytes, and its expression increases with age, we generated hepatocyte-specific MLKL-overexpressing mice (MLKLHepOE) to determine its role in liver aging. Unexpectedly, MLKL overexpression in hepatocytes did not induce necroptosis, but instead upregulated markers of cellular senescence (cell cycle arrest genes and SASP factors), increased macrophage infiltration, and elevated M1 macrophage marker expression. Electron microscopy and mitochondrial analyses revealed abnormal mitochondrial morphology, elevated oxidative stress, and disrupted mitochondrial dynamics, while lipidomics demonstrated alterations in hepatic lipid metabolites. In agreement with our observations in MLKLHepOE mice, MLKL overexpression in AML12 hepatocytes impaired mitochondrial respiration, increased proinflammatory extracellular vesicle (EV) release, and induced senescence markers, without triggering cell death. Together, these findings reveal a non-lethal, non-necroptotic role for MLKL in promoting hepatocyte senescence and metabolic dysfunction via mitochondrial impairment and EV-mediated inflammation. Our study highlights MLKL as a novel driver of liver inflammaging and a potential therapeutic target for age-related liver disease.

cell biology↗

bioRxiv2025

Predicting RNA 3D structure and conformers using a pre-trained secondary structure model and structure-aware attention

Determining RNA 3D structure and conformers remains a grand challenge in structural biology. In this work, we propose trRosettaRNA2, a deep learning-based end-to-end approach to this problem. Considering the scarcity of RNA 3D structure data, trRosettaRNA2 integrates an auxiliary secondary structure (SS) prior module, pre-trained on extensive SS data, to generate informative base-pairing priors. This module also serves as an independent RNA secondary structure prediction method, trRNA2-SS, and achieves state-of-the-art performance. To enable end-to-end prediction, trRosettaRNA2 employs SS-aware attention to generate RNA 3D structure and conformers. Rigorous benchmarks demonstrate that trRosettaRNA2 outperforms other RNA 3D structure prediction methods, despite using significantly fewer parameters and computational resources. Notably, its flexibility in leveraging diverse secondary structure inputs provides a pathway to generate accurate 3D structure and explore the RNA conformers. Based on trRosettaRNA2, our group, Yang-Server, ranks as the top automated server for RNA structure prediction in the CASP16 blind test, surpassing AlphaFold3. This performance highlights that trRosettaRNA2 represents a solid step towards the "AlphaFold moment" for RNA structure prediction. Application to the RNase P RNA demonstrates that trRosettaRNA2 successfully captures its structural heterogeneity even without requiring experimental data, showing its potential to predict RNA conformational ensembles.

bioinformatics↗