Nature Portfolio

Abstract Proteins are essential to life, and understanding their structure can facilitate a mechanistic understanding of their function. Through an enormous experimental effort 1–4 , the structures of around 100,000 unique proteins have been determined 5 , but this represents a small fraction of the billions of known protein sequences 6,7 . Structural coverage is bottlenecked by the months to years of painstaking effort required to determine a single protein structure. Accurate computational approaches are needed to address this gap and to enable large-scale structural bioinformatics. Predicting the three-dimensional structure that a protein will adopt based solely on its amino acid sequence—the structure prediction component of the ‘protein folding problem’ 8 —has been an important open research problem for more than 50 years 9 . Despite recent progress 10–14 , existing methods fall far short of atomic accuracy, especially when no homologous structure is available. Here we provide the first computational method that can regularly predict protein structures with atomic accuracy even in cases in which no similar structure is known. We validated an entirely redesigned version of our neural network-based model, AlphaFold, in the challenging 14th Critical Assessment of protein Structure Prediction (CASP14) 15 , demonstrating accuracy competitive with experimental structures in a majority of cases and greatly outperforming other methods. Underpinning the latest version of AlphaFold is a novel machine learning approach that incorporates physical and biological knowledge about protein structure, leveraging multi-sequence alignments, into the design of the deep learning algorithm.

Abstract Livestock manure is a prominent One Health setting for antimicrobial resistance (AMR), yet the contribution of viruses to resistome assembly remains unresolved. Here we compile the Livestock Manure Virus Database (LMVD), a global catalog of 1,330,372 viral operational taxonomic units reconstructed from 4,017 metagenomes. Livestock manure viral communities exhibit restricted biogeographic distributions, and antibiotic use and farming scale show the highest relative predictive importance for viral diversity. Viral genomes contained 78 antibiotic resistance gene (ARG) subtypes, representing 9.1% of resistome diversity in livestock manure, but virus-encoded ARGs were infrequent. Bacteriophages instead showed extensive predicted links to bacterial hosts, with nearly all antibiotic-resistant bacteria linked to viruses, predominantly temperate viruses, and enrichment in pathogenic antibiotic-resistant bacteria. Together, these findings place bounds on the contribution of viruses to direct ARG dissemination and indicate that lysogenic interactions with resistant hosts may shape AMR risk.

Abstract Enterovirus D68 (EV-D68) is a common cause of mild respiratory illness in children but is also linked to acute flaccid myelitis (AFM), a rare and severe neurological condition which can result in permanent paralysis. Tracking EV-D68 transmission is difficult because symptoms overlap with other common childhood illnesses, making case reporting unreliable. As a result, the burden of EV-D68 transmission remains poorly understood, particularly prior to the first widely recognised outbreak in 2014. Phylodynamic approaches offer a powerful alternative to case-based surveillance, inferring epidemiological parameters directly from pathogen genetic sequences. Here we show, using a birth-death-sampling phylodynamic model applied to publicly available EV-D68 sequences from the 1990s to 2023 collected in Europe and North America, that seasonal EV-D68 transmission was well established long before the 2014 outbreak brought the virus to widespread attention. Furthermore, we show that AFM case burden is strongly associated with the transmission dynamics of a specific EV-D68 clade, clade B, an association not seen for clades A and D. These findings suggest that clade-specific biological differences drive AFM risk, with important implications for surveillance prioritisation and the timing of future burden of neurological disease.

Abstract Effective pulpectomy of maxillary primary molars depends on a clear understanding of their complex internal root canal anatomy, yet high-resolution data on the second mesiobuccal (MB2) canal in the primary dentition remain scarce. Using micro-computed tomography (micro-CT; SkyScan 1172; 13.61 μm isotropic voxel size), the three-dimensional root and canal morphology of 60 extracted maxillary primary molars were charaterized, focusing on the prevalence, location, and morphometry of the MB2 canal, along with inter-orifice distances, canal diameters, and dentin wall thickness. Three-rooted molars predominated (83.3%). An MB2 canal was present in 43.3% of mesiobuccal roots and, when present, was consistently located palatal to the main mesiobuccal (MB1) canal, originating predominantly in the coronal third. Long-oval and oval cross-sections together accounted for almost all canals. The MB1–MB2 inter-orifice distance was the shortest of any canal pair (mean 1.12 mm), and buccolingual diameters exceeded mesiodistal diameters in the mesiobuccal, MB2, and distobuccal roots but not the palatal root. Apical dentin was thinnest in the palatal roots (0.10 mm). Maxillary primary molars therefore combine a frequently overlooked MB2 canal with thin apical walls. These findings may help clinicians plan access cavities, select instruments, and perform conservative apical preparation during pulpectomy procedures.

Abstract Corneal diseases are among the leading causes of blindness, and transplantation may be the only treatment option. Studies providing a comprehensive economic analysis of the corneal tissue processing chain, from procurement to release for transplantation by eye banks, are scarce. This study aimed to estimate the cost of preparing donated corneal tissue for transplantation at an Eye Bank. We performed a cost analysis using the Time-Driven Activity-Based Costing (TDABC) method based on data from the stages of corneal tissue preparation. The study population consisted of data from corneal donors collected in 2023. The median cost per donor was estimated at USD 454.65. Supplies (43%), human resources (32%), and operational costs (23%) were the costliest components. Multiple regression analysis identified predictors influencing costs. This study highlighted critical stages in the process and resource consumption. Investing in domestic pharmaceutical production, improving interinstitutional communication, training healthcare professionals, and providing logistical support may reduce costs and contribute to the management and sustainability of Eye Banks.

Abstract Accurate detection and depth perception of pavement targets from close-range remote sensing observations are important for intelligent road monitoring and transportation infrastructure assessment. However, existing stereo matching methods often produce unstable disparity estimates in weak-texture and low-height pavement regions, such as manhole covers and speed bumps, where the disparity distribution is easily affected by matching ambiguity and image noise. To address this issue, a detection-guided disparity post-processing strategy named Detection-Aware Disparity Refinement (DADR) is proposed. A custom pavement dataset containing approximately 3000 annotated images is first constructed from a binocular vision-based remote sensing platform, covering five pavement target categories, including manhole covers, speed bumps, road damage, storm drains, and iron plates. YOLOv11 is employed to detect target regions, while FoundationStereo is used to generate initial disparity maps from calibrated stereo images. Instead of modifying the stereo matching network, the proposed method performs a rule-based post-processing operation that combines detection bounding boxes, a local ground-reference disparity, and confidence-weighted fusion to stabilize the disparity representation within detected target regions. Experimental results show that the YOLOv11 detector achieves a Precision of 0.89, a Recall of 0.88, and an mAP@50 of 0.895, and compares favourably with other one-stage detectors trained under the same protocol. Visual analysis with magnified regions of interest shows that the proposed strategy produces more spatially coherent disparity responses for weak-texture pavement targets, and an ablation study identifies the local ground reference and the adaptive confidence weighting as the main contributing components. Quantitative evaluation indicates that the within-region disparity standard deviation decreases from 3.81 px to 1.25 px after refinement. Because no independent depth ground truth is available, this result is reported as improved local disparity consistency rather than improved absolute depth accuracy. The proposed approach offers a lightweight, detector-agnostic disparity stabilization strategy for low-height pavement targets in close-range remote sensing scenarios and provides useful geometric cues for subsequent road-condition assessment, while its reliance on a locally planar ground reference and on detector reliability is discussed as a limitation.