Research on plant viruses depends on taxonomy, nucleic acid sequences, host and disease records, molecular targets, and antiviral assays. These data are distributed across separate resources, making integrated retrieval and comparison difficult. PlantVirDisConDB (https://dpai.ccnu.edu.cn/PVDCDB) is freely accessible without login or registration and links these data through curated plant virus records. At the July 2026 reporting date, it comprised, in round numbers, 5000 virus and subviral agent records and 38 000 sequence accessions. It also included approximately1900 hosts, 9000 virus–host associations, 4800 disease names, 5000 structure records, and 25 000 small molecules. Direct assay records for plant viruses are distinguished from general antiviral background data; experimental and predicted structures, and measured activities and model outputs, are also labelled separately. Linked record pages support retrieval across these data types, access to individual sequence records, target searches, and molecular similarity searches. An activity prediction module ranks candidate molecule–virus pairs with applicability warnings and does not imply experimental or field efficacy. PlantVirDisConDB links viral nucleic acid entities to hosts, phenotypes, targets, and antiviral information and is updated through continuing curation and user feedback.

Small self-cleaving ribozymes are catalytic RNAs discovered in viroid-like agents, which are replicating circular RNAs (circRNAs) postulated as relics of a prebiotic RNA world. In the last decade, however, small ribozymes have also been detected across the tree of life, from bacterial to human genomes, and more recently, in unusual circRNA viruses. Here we report the conserved occurrence of diverse small ribozymes within the linear genomes of typical single- and double-stranded RNA viruses infecting fungi and plants. We identified type I hammerhead ribozymes in the 5′-UTRs of chrysovirids and fusarivirids, displaying self-cleaving activity in vitro and in vivo. Similar hammerhead, as well as hepatitis delta and twister ribozymes, occur in diverse megabirna-, hypo-, fusagra-, toti-, or tombus-like viruses. The ribozymes appear as isolated motifs within the UTRs or as tandem pairs that encompass small RNA segments (186–399 nt) resembling Zetavirus-like sequences. In vivo characterization of the 5′-UTR from a fungal chrysovirus revealed that its RNA self-cleaving activity is essential for protein translation initiation. Analogous experiments in plants with diverse ribozyme motifs confirmed that self-cleaving activity alone can induce cap-independent translation. We conclude that RNA self-cleaving activity, historically linked to the rolling circle replication of viroid-like circRNAs, appears to be co-opted by some linear RNA viruses for translational roles.

Prebiotics and fermentable fibre modulate gut microbiota composition and function in ways considered beneficial to health. However, both microbiota and clinical responses can vary widely among individuals, which reduces their efficacy as dietary interventions across populations and thus represents a key challenge for the field. In this Review, we discuss potential reasons for the individualized responses to prebiotics and putative prebiotic fibres and how this variability influences interactions among gut microbial species as well as between microorganisms and fibre structures. We further describe strategies for predicting and optimizing responses to prebiotic and putative prebiotic fibre treatments. Specifically, we highlight methods for targeted and predictable individual responses using systems biology, mechanistic modelling and machine learning approaches. We also consider how the physicochemical structures of fermentable fibre influence similarity in gut microbial responses. By integrating this knowledge, solutions are proposed to overcome inter-individual variability and improve the clinical efficacy and utility of prebiotic fibres. Prebiotics and dietary fibre show highly individualized effects on gut microbiota and clinical outcomes, challenging their consistent effectiveness across populations. This Review examines the drivers of this variability, presents predictive approaches and discusses strategies to optimize prebiotic responses across individuals.

Chemical modifications of transfer RNAs (tRNAs) are crucial for tRNA stability, codon decoding, translational fidelity, and tRNA-derived small RNA (tsRNA) biogenesis. Changes in these modifications are associated with stress responses, physiological regulation, and human disease, underscoring the need to integrate modification information with experimental evidence and biological context. We previously established tModBase, a web resource to organize and explore tRNA modification-related data. Its first release integrated modification sites, high-throughput sequencing-derived signals, and condition-dependent changes. Since then, tRNA profiling has expanded through modification-specific and direct RNA sequencing strategies, with rapidly growing datasets across species and conditions. Here, we present tModBase v2.0 (http://www.tmodbase2.com/), containing 528 datasets across 23 species, 40 sequencing and library-preparation methods, and 42 modification types. It includes 68 third-generation and 210 modification-specific sequencing datasets, extending coverage of direct-RNA and modification-specific approaches. tModBase v2.0 further integrates sequencing-derived signals, condition-dependent information, and biological annotations linking modifications with enzymes, genetic variants, tsRNAs, diseases, and expression profiles. The LLM- and retrieval-augmented generation-powered assistant tModBot enables natural-language exploration across database modules while grounding responses in retrieved tModBase evidence. tModBase v2.0 therefore provides an expanded resource for investigating the diversity, dynamics, and biological context of tRNA modifications.

During translation, transfer RNAs (tRNAs) deliver specific amino acids to the ribosome in a coordinated manner with the sequence encoded by the messenger RNA. Despite their central role in protein synthesis, their precise contribution to the modulation of translation remains poorly understood, primarily due to lack of methods to characterize tRNA abundances and their modifications from actively translating ribosomes. Here we develop tRIBO-seq, a simple and robust nanopore-based method to selectively capture ribosome-associated native tRNA populations (ribo-tRNAs) from actively translating ribosomes, providing tRNA abundance, modification and fragmentation information from a single experiment. Using tRIBO-seq, we find that ribo-tRNAs, but often not otal tRNAs, are significantly altered on stress. Notably, we find that tRNAome alterations strongly vary depending on the stress type; while viral infection and leucine deprivation leads to changes in ribo-tRNA abundances, methionine starvation causes a dramatic loss of methyl-based tRNA modifications. By contrast, we find that arsenite exposure does not alter tRNA abundances nor modification patterns, but rather causes major fragmentation events in selected subsets of ribo-tRNAs. Altogether, tRIBO-seq offers a robust and reproducible approach to map the full tRNA landscape—capturing tRNA abundance, modification and fragmentation patterns—across both total and actively translating tRNA populations, revealing the dynamics of the tRNAome with unprecedented resolution, in a single experiment. tRIBO-seq distinguishes between ribosome-associated tRNAs and total tRNAs using nanopore sequencing, thus revealing their potentially distinct abundance, modification and fragmentation patterns in various conditions.

Predicting the functional effects of missense variants is limited both by predictor performance and by the lack of principled strategies to integrate heterogeneous multiplexed assays of variant effect (MAVEs), including deep mutational scanning, into transferable sequence-function models. Here, we show that redundancy among partially overlapping deep mutational scanning experiments, often treated as noise, yields a consensus landscape of mutational constraint. We introduce “variant soundness”, an overlap-aware rank-fusion score that uses cross-experiment consistency to derive an assay-agnostic ordering of mutational tolerance within proteins. Applied to ~1100 MAVEs spanning over two million missense variants, this approach reveals a mutational landscape enriched for structural stability constraints. Guided by this integrated signal, we develop ESMRank, a sequence-based learning-to-rank predictor combining ESM2 representations with physicochemical descriptors. Under strict protein-level partitioning, ESMRank outperforms widely used predictors across abundance-, folding-, and stability-related assays in ProteinGym and Domainome. Predicted deleterious variants are enriched for ClinVar pathogenicity and, in CFTR, correlate with folding efficiency, channel activity, and pharmacological responsiveness. Together, these results turn assay redundancy into a scalable resource for interpretable missense variant prediction.

RNA-targeted small-molecule discovery is challenged by the scarcity of experimentally characterized RNA–ligand complexes and by the limited transferability of current computational approaches across chemically diverse systems. Here we present P-RISE, a physics-grounded framework for RNA–ligand binding prediction, and use it to identify the physical determinants underlying reliable performance. Systematic evaluation of force fields, ionic environments, dielectric treatments, and entropy corrections reveals three coupled requirements: preservation of RNA-binding pocket integrity under ion-sensitive conditions; an effective electrostatic error cancellation regime enabled by an elevated internal dielectric constant; and an empirically observed coupling between entropy correction and solvent modeling. Integrating these features, P-RISE provides stable affinity ranking, physics-grounded delineation of its physicochemical applicability domain, accurate metal ion site localization, and improved binding pose discrimination relative to established RNA docking and scoring approaches. Crucially, benchmarking against AlphaFold3 reveals that co-folding models, optimized for structure generation rather than energetic ranking, benefit from an explicit physics-based evaluation layer. Furthermore, diverse machine-learning comparisons demonstrate that purely data-driven methods lack stability and generalization in this data-poor regime. Ultimately, P-RISE provides a mechanistic understanding of RNA–ligand modeling, serving as a principled physical baseline that defines reliable prediction boundaries, and guides future data-driven integration.

The circadian clock orchestrates daily physiological processes through coordinated regulation of clock-controlled gene expression across multiple molecular layers. However, systematic characterization of this multilayered rhythmic regulation remains challenging because relevant information is distributed across independent datasets and rhythmicity is evaluated using diverse analytical approaches. Here, we developed CircaVal, a gene-centric multi-omics platform for the integrated exploration and interpretation of circadian regulatory features in humans, mice, and baboons. To support this framework, CircaVal processes public datasets using standardized, data-type-specific pipelines, and gene rhythmicity is integrated with tissue-specific expression, cross-species orthology, functional annotations, and multi-omics regulatory features. CircaVal further introduces an integrated rhythmicity score (IRS), which combines concordance across multiple rhythm-detection methods with reproducibility across independent datasets to provide a quantitative basis for prioritizing candidate rhythmic genes. The platform enables users to examine gene rhythmicity across datasets, identify tissue-enriched rhythmic genes, compare rhythmicity across species, and explore associated regulatory features. Additional functions include interactive gene-centric exploration, genome-browser visualization, rhythmicity analysis of user-uploaded time-series data, and programmatic data access. Together, these capabilities provide an integrated framework for rhythmic gene prioritization and gene-level interpretation of circadian regulatory features across tissues, species, datasets, and omics data types. CircaVal is freely available at https://CircaVal.com.

Genome expansion and contraction are driven by transposable element activity, but the underlying dynamics remain enigmatic due to a lack of historical records tracing these changes. Here, we present PrinTE, a forward-time simulator of whole-genome sequences with customizable transposon dynamics for identifying the evolutionary parameters most consistent with an observed genome. Across simulations spanning four orders of magnitude, PrinTE recovers transposon insertion and deletion rates with less than 6% error. We demonstrate PrinTE on Arabidopsis and the fungal subdivision Pucciniomycotina, including myrtle rust (Austropuccinia psidii), which expanded its genome size to 1,018 Mb through transposon amplifications. Our simulations show that recombination-based removal of LTR retrotransposons alone cannot prevent genome obesity, and that preventing their invasion in the first place is more effective. We hypothesize that A. psidii might have experienced a prolonged expansion phase followed by a potent and ongoing period of contraction. The authors present PrinTE, a freely available simulator that evolves genomes forward in time to reveal how transposable elements drive genome expansion and contraction, and use it to test hypotheses about genome size evolution in fungi and plants.