Biology subjects

Kihara, D.

Publications and source records attributed to Kihara, D..

At least 19 recordsLinked to original sources

bioRxiv2026

NuFold RNA Structure Database: Large-Scale RNA Tertiary Structure Prediction with Confidence Metrics

Summary: To bridge the gap between rapid RNA sequence discovery and slow experimental characterization, we present the NuFold RNA Structure Database (https://nufold.kiharalab.org/), a resource providing large-scale, precomputed full-atom RNA tertiary structures. By removing the computational barriers associated with running the method locally, this database makes high-accuracy RNA structure models readily accessible to the biological research community for a wide range of applications. Availability and implementation: All predicted structures and corresponding data are available at https://nufold.kiharalab.org/. The source code for the NuFold program is available at https://github.com/kiharalab/NuFold. Supplementary Information: For more insight into resulting statistics gathered for this project, see https://nufold.kiharalab.org/statistics. For more insight into how to utilize the webserver itself, see https://nufold.kiharalab.org/faq and https://nufold.kiharalab.org/tutorial.

bioinformatics↗

bioRxiv2026

Distance-AF Multimer: A Multi-Step Approach for Protein Complex Structure Modeling with User-Defined Distance Constraints

The three-dimensional structure of proteins provides essential insight into the molecular mechanisms underlying their function. Many biological activities are mediated through protein interactions, making accurate structural determination of protein complexes important. Despite significant advances in computational protein structure prediction, there are still many cases where protein complexes are not correctly assembled. In particular, there are cases where experimental data, such as distances between residue pairs in interacting proteins are available, yet existing prediction methods fail to produce models consistent with these data. Here, we present Distance-AF Multimer, a method that guides and refines AlphaFold-Multimer docking predictions using user-provided residue pair distance information. This approach extends our previously successful data-guided single-protein modeling method, Distance-AF. Distance-AF Multimer operates in a two-step process, starting from multiple initial conformations and iteratively refining them to satisfy the provided distance constraints. We evaluated Distance-AF Multimer on protein-complex benchmarks spanning different chain counts and docking errors, demonstrating competitive or superior performance compared to other existing deep-learning-based structure prediction methods. In addition to dimer cases, we also demonstrated the potential of Distance-AF Multimer for large-scale protein assemblies containing multiple chains, including those with more than 5 chains.

bioinformatics↗

bioRxiv2026

PathFold: Predicting the Entire Protein Folding Pathway from Protein Sequence Alone

Recent advances in protein structure prediction, exemplified by AlphaFold, have largely addressed the determination of static structures, one aspect of the protein folding problem. However, predicting folding pathways, by which proteins reach their native states, remains a significant challenge. Here, we present PathFold, a deep learning framework that predicts protein folding pathways directly from sequence information. PathFold leverages an AlphaFold-based module to extract structural information from the sequence and generates a progressive folding trajectory from an extended conformation using a diffusion model. By modeling the full trajectory, it enables prediction of folding intermediates and transition pathways, analogous to those observed in steered molecular dynamics (SMD) simulations. The predicted pathways reveal well-defined intermediates and sequential folding events, and show agreement with experimental folding data, including measured {Phi}-values.

bioinformatics↗

bioRxiv2026

Prot-LAMBDA: Explicit Distance Learning Enhances Structural Reasoning in Protein Language Models

Protein language models (PLMs) learn evolutionary information from large-scale sequence data, but three-dimensional relationships are encoded only implicitly. Here, we introduce Prot-LAMBDA (Protein LAnguage Model Boosted with Distance Awareness), a PLM that explicitly incorporates spatial relationships by coupling residue embeddings with inter-residue contacts. Prot-LAMBDA improves performance across diverse structure-related tasks, including contact, secondary structure, backbone geometry, solvent accessibility, and protein fold prediction. Notably, it achieves a twofold improvement in long-range contact recall and an 11.7% reduction in {psi}-angle prediction error relative to ESM2-3B. Despite having approximately fivefold fewer parameters, Prot-LAMBDA also improves 3D structure prediction over ESM2-3B by 5-7% in TM-score when coupled to the same structure-prediction module. Building on these representations, we developed LambdaFold, a lightweight distance-guided structure prediction framework that achieves performance comparable to ESMFold on proteins strictly non-redundant to the training data. Finally, retrieval-augmented integration of structural templates increases mean TM-score substantially for targets with high template coverage and rescues several incorrect folds. Together, these results demonstrate that explicit spatial constraints enable efficient and generalizable structural representation learning and protein structure prediction.

bioinformatics↗

bioRxiv2026

Structural determination of human nucleosomes reconstituted by the ExACT platform

The structure and function of eukaryotic chromatin have been extensively studied using conventional salt dialysis-based nucleosome assembly, which has provided fundamental structural insights into nucleosomes as well as a mechanistic understanding of their roles in DNA replication, repair and gene expression. Recently, we developed a labor-saving and time-efficient nucleosome assembly method using a wheat germ cell-free Expression and Assembly Coupled Technology (ExACT), which is emerging as a powerful tool for chromatin research. Here we report cryo-electron microscopy (cryo-EM) structures of human H3.1- and H3.3-containing nucleosomes assembled using this approach and validated by deep-learning-based amino acid-wise model quality (DAQ) scoring. The structures of H3.1- and H3.3-nucleosomes are nearly identical to previously reported models, confirming the structural fidelity of the method. In addition, we determined the previously unreported structure of the primate-specific H3.X-containing nucleosome. Together, these findings validate the cell-free co-expression nucleosome assembly platform and establish this method as a robust framework for biochemically investigating chromatin dynamics.

molecular biology↗

bioRxiv2026

DMcloud: Macromolecular Structure Modeling Using Local Structure Fitting for Medium to Low Resolution cryo-EM maps

Cryogenic electron microscopy (cryo-EM) has become an essential experimental approach in structural biology for determining macromolecular structures. When the resolution of a cryo-EM map is worse than approximately 5 [A], fitting known or predicted molecular models into the map becomes a common strategy for interpretation. However, accurately fitting biomolecular models into cryo-EM maps, particularly for large macromolecular complexes, remains challenging when the input structure models contain errors or are in a conformation different from that represented in the map. Here, we present DMcloud, a method for local structure fitting of proteins and nucleic acids in cryo-EM maps. Instead of forcing an entire input model into the map, DMcloud divides input structures into local regions, identifies regions that are supported by the density, removes unsupported regions, and assembles the retained regions into a final model. We benchmarked DMcloud on 176 cryo-EM maps, including intermediate and high-resolution maps that include proteins, DNAs, or RNAs. For EM maps in the 5.0-10.0 [A] and 2.5-5.0 [A] resolution ranges, DMcloud achieved average sequence modeling coverage of 0.49 and 0.70, respectively. For DNA/RNA maps, DMcloud achieved an average sequence coverage of 0.75. Across all datasets, DMcloud consistently outperformed existing methods in model accuracy, map-model correlation, and modeling coverage.

bioinformatics↗

bioRxiv2026

DAQplugin: Deep Learning based Real-time Model Evaluation Plugin for ChimeraX

Although an increasing number of protein structures are determined by cryogenic electron microscopy (cryo-EM), structure modeling frequently suffers from residue misassignments and sequence register shifts, particularly in regions with ambiguous density. Here, we present DAQplugin, a ChimeraX plugin for real-time evaluation of protein models against cryo-EM density maps using the deep-learning-based residue-wise model quality (DAQ) score. Unlike existing validation tools that are typically applied after model construction, DAQplugin enables interactive validation during model building and refinement. DAQ has been shown to accurately identify residue assignment errors, including sequence register shifts, as well as local conformational modeling errors. DAQplugin also provides guidance for correcting sequence register shifts by suggesting alternative residue placements along the backbone. The plugin is computationally efficient and runs on standard CPUs without requiring GPU hardware, enabling deep-learning-based validation on ordinary laptops during interactive model building, model-map fitting, and refinement. DAQplugin facilitates more accurate interpretation of cryo-EM density maps and improve the reliability assessment of protein structure models.

bioinformatics↗

bioRxiv2026

Direct Detection and Atomic Modeling of Ligands in Cryo-EM Maps Using Deep Learning

Cryogenic electron microscopy (cryo-EM) has become an increasingly important for structure-based drug discovery by enabling characterization of interactions between macromolecules and small-molecule ligands. However, computational interpretation of ligand density remains challenging, particularly when ligand locations are unknown or local map resolution is limited. Existing methods generally require well-resolved macromolecular structures and predefined binding sites, limiting their applicability during early-stage structure determination. To date, no approach has been able to both reliably detect ligand density and subsequently reconstruct ligand atomic structures directly from experimental cryo-EM maps. Here, we present Emap2lig, a two-stage deep learning framework for automated ligand detection and atomic modeling directly from cryo-EM maps. Emap2lig-Find identifies ligand-associated densities and remains effective for maps at resolutions as low as [~]5 [A]. Emap2lig-Build subsequently uses a diffusion-based generative model to build atomic ligand structures. Together, Emap2lig provides a unified framework for ligand discovery and modeling across a broad range of resolutions.

bioinformatics↗

bioRxiv2026

Accurate Macromolecular Complex Modeling for Cryo-EM with CryoZeta

Cryogenic electron microscopy (cryo-EM) has become a widely used technique for determining the three-dimensional structures of biological macromolecules. Despite its advantages, building accurate structural models from cryo-EM data remains challenging, particularly at non-atomic resolutions. Here, we present CryoZeta, a de novo structure modeling program that leverages a diffusion-based generative deep neural network to integrate cryo-EM map density features with a biomolecular structure prediction pipeline similar to Alphafold3. By jointly leveraging sequence information and density-based features, CryoZeta generates highly accurate structural models that are consistent with the experimental map density. Evaluated on benchmark datasets covering protein complexes, protein-nucleic acid assemblies, and nucleic acid-only systems at resolutions up to 10 [A], CryoZeta consistently outperforms existing cryo-EM modeling methods in atomic accuracy. These results highlight the benefits of directly incorporating cryo-EM density into modern structure prediction pipelines and establish the method as a robust tool for automated, high-fidelity modeling from cryo-EM maps.

bioinformatics↗

bioRxiv2025

Queryome: Orchestrating Retrieval, Reasoning, and Synthesis across Biomedical Literature

The rapid expansion of biomedical literature has made comprehensive manual synthesis increasingly difficult to perform effectively, creating a pressing need for AI systems capable of reasoning across verified evidence rather than merely retrieving it. However, existing retrieval-augmented generation (RAG) methods often fall short when faced with complex biomedical questions that require iterative reasoning and multi-step synthesis. Here, we developed Queryome, a deep research system consisting of specialized large language model (LLM) agents that can adapt their orchestration dynamically to a wide range of queries. Using a hybrid semantic-lexical retrieval engine spanning 28.3 million PubMed abstracts, it performs iterative, evidence-grounded synthesis. On the MIRAGE benchmark, Queryome achieved 88.98 % accuracy, surpassing prior systems by up to 14 points, and improved reasoning accuracy on the biomedical Humans Last Exam (HLE) subset from 15.8% to 19.3%. Moreover, in a task for constructing a review article, it earned the highest composite score in comparison with Deep Research from OpenAI, Google, Perplexity, and Scite.AI, reflecting its strong literature retrieval and synthesis capabilities.

bioinformatics↗

bioRxiv2025

Structure of bacteriophage P1 head provides insight into capsid polymorphism

The highly regulated assembly of different proteins forming mature P1 myophage capsid show a unique polymorphism where the virion population is primarily divided in brackets of two different capsid sizes. The contrasting capsids, sized 95nm (capsid_L), triangulation number (T)=13 and 68nm (capsid_S) with T=7, both in Dextro format and have the same protein forming the phage head with similar intra and inter capsomeric interactions. Comparative study of the electron density maps of the capsids reveals the presence of protein appendages DarA and Hdf below the 5-fold symmetric region inside capsid_L that anchor the phage dsDNA to the capsid-shell. Such densities are also noticed in large capsids depleted off their dsDNA but not observed in capsid_S. Deficiency of these appendages in capsid_S results in a smaller sized dsDNA packed inside capsid_S. Despite missing out on nearly 60% of the phage dsDNA virions with capsid_S has all the essential genes responsible for the formation of a fully mature thermodynamically stable virion particle with structurally identical tail and baseplate as of capsid_L. Co-existence of both stable conformation in a single lysate is a unique phenomenon in the phage community and suggest that DarA and Hdf play an important role in phage capsid morphogenesis.

biophysics↗

bioRxiv2025

Blind Prediction of Complex Water and Ion Ensembles Around RNA in CASP16

Biomolecules rely on water and ions for stable folding, but these interactions are often transient, dynamic, or disordered and thus hidden from experiments and evaluation challenges that represent biomolecules as single, ordered structures. Here, we compare blindly predicted ensembles of water and ion structure to the cryo-EM densities observed around the Tetrahymena ribozyme at 2.2-2.3 [A] resolution, collected through target R1260 in the CASP16 competition. 26 groups participated in this solvation cryo-ensemble prediction challenge, submitting over 350 million atoms in total, offering the first opportunity to compare blind predictions of dynamic solvent shell ensembles to cryo-EM density. Predicted atomic ensembles were converted to density through local alignment and these densities were compared to the cryo-EM densities using Pearson correlation, Spearman correlation, mutual information, and precision-recall curves. These predictions show that an ensemble representation is able to capture information of transient or dynamic water and ions better than traditional atomic models, but there remains a large accuracy gap to the performance ceiling set by experimental uncertainty. Overall, molecular dynamics approaches best matched the cryo-EM density, with blind predictions from bussilab_plain_md, SoutheRNA, bussilab_replex, coogs2, and coogs3 outperforming the baseline molecular dynamics prediction. This study indicates that simulations of water and ions can be quantitatively evaluated with cryo-EM maps. We propose that further community-wide blind challenges can drive and evaluate progress in modeling water, ions and other previously hidden components of biomolecular systems.

biophysics↗

bioRxiv2025

CryoDataBot: a pipeline to curate cryoEM datasets for AI-driven structural biology

Cryogenic electron microscopy (cryoEM) has revolutionized structural biology by enabling atomic-resolution visualization of biomacromolecules. To automate atomic model building from cryoEM maps, artificial intelligence (AI) methods have emerged as powerful tools. Although high-quality, task-specific datasets play a critical role in AI-based modeling, assembling such resources often requires considerable effort and domain expertise. We present CryoDataBot, an automated pipeline that addresses this gap. It streamlines data retrieval, preprocessing, and labeling, with fine-grained quality control and flexible customization, enabling efficient generation of robust datasets. CryoDataBots effectiveness is demonstrated through improved training efficiency in U-Net models and rapid, effective retraining of CryoREAD, a widely used RNA modeling tool. By simplifying the workflow and offering customizable quality control, CryoDataBot enables researchers to easily tailor dataset construction to the specific objectives of their models, while ensuring high data quality and reducing manual workload. This flexibility supports a wide range of applications in AI-driven structural biology.

molecular biology↗

bioRxiv2025

Divergence of an extracellular contractile injection system infectivity elucidated by high resolution structural studies of its tail-baseplate complex

Divergence to the infectivity practice by extracellular contractile injection systems (eCISs) is displayed by myophage P1 in its lytic phase involving its baseplate receding away from the host bacterium during infection. Atomic structure of the proteins forming P1s Tail Baseplate (TB) complex, determined here, using cryo electron microscopy are employed to identify a sequence of viral events explaining this unique phenomenon. The P1 baseplate is found to be devoid of protein appendages that are necessary for anchoring it to the host bacterium. To compensate this deficiency, P1s Long Tail Fibers (LTFs) affix to the hosts exterior, straighten up imparting stability to the virion for pursuing the infection process, thereby eliciting the baseplate hub and the tail sheath to ascend away, resulting in uniform compression of the latter. Upon the maximum unkinking of LTFs, a descending corkscrew motion commences that results in the non-uniform downward compression of the tail sheath, the tail tube and baseplate needle that perforates through the hosts membrane-cytoplasm, leading to a successful phage-bacterium infection.

biophysics↗

bioRxiv2024

Proteomic changes orchestrate metabolic acclimation of a unicellular diazotrophic cyanobacterium during light-dark cycle and nitrogen fixation states

Cyanobacteria have developed an impressive array of proteins and pathways, each tailored for specific metabolic attributes, to execute photosynthesis and biological nitrogen (N2)-fixation. An understanding of these biologically incompatible processes provides important insights into how they can be optimized for renewable energy. To expand upon our current knowledge, we performed label-free quantitative proteomic analysis of the unicellular diazotrophic cyanobacterium Crocosphaera subtropica ATCC 51142 grown with and without nitrate under 12-hour light-dark cycles. Results showed significant shift in metabolic activities including photosynthesis, respiration, biological nitrogen fixation (BNF), and proteostasis to different growth conditions. We identified 14 nitrogenase enzymes which were among the most highly expressed proteins in the dark under nitrogen-fixing conditions, emphasizing their importance in BNF. Nitrogenase enzymes were not expressed under non nitrogen fixing conditions, suggesting a regulatory mechanism based on nitrogen availability. The synthesis of key respiratory enzymes and uptake hydrogenase (HupSL) synchronized with the synthesis of nitrogenase indicating a coordinated regulation of processes involved in energy production and BNF. Data suggests alternative pathways that cells utilize, such as oxidative pentose phosphate (OPP) and 2-oxoglutarate (2-OG) pathways, to produce ATP and support bioenergetic BNF. Data also indicates the important role of uptake hydrogenase for the removal of O2 to support BNF. Overall, this study expands upon our knowledge regarding molecular responses of Crocosphaera 51142 to nitrogen and light-dark phases, shedding light on potential applications and optimization for renewable energy.

microbiology↗

bioRxiv2024

Proteomic analysis of unicellular cyanobacterium Crocosphaera subtropica ATCC 51142 under extended light or dark growth

The daily light-dark cycle is a recurrent and predictable environmental phenomenon to which many organisms, including cyanobacteria, have evolved to adapt. Understanding how cyanobacteria alter their metabolic attributes in response to subjective light or dark growth may provide key features for developing strains with improved photosynthetic efficiency and applications in enhanced carbon sequestration and renewable energy. Here, we undertook a label-free proteomic approach to investigate the effect of extended light (LL) or extended dark (DD) conditions on the unicellular cyanobacterium Crocosphaera subtropica ATCC 51142. We quantified 2287 proteins, of which 603 proteins were significantly different between the two growth conditions. These proteins represent several biological processes, including photosynthetic electron transport, carbon fixation, stress responses, translation, and protein degradation. One significant observation is the regulation of over two dozen proteases, including ATP dependent Clp-proteases (endopeptidases) and metalloproteases, the majority of which were upregulated in LL compared to DD. This suggests that proteases play a crucial role in the regulation and maintenance of photosynthesis, especially the PSI and PSII components. The higher protease activity in LL indicates a need for more frequent degradation and repair of certain photosynthetic components, highlighting the dynamic nature of protein turnover and quality control mechanisms in response to prolonged light exposure. The results enhance our understanding of how Crocosphaera subtropica ATCC51142 adjusts its molecular machinery in response to extended light or dark growth conditions. O_FIG O_LINKSMALLFIG WIDTH=200 HEIGHT=180 SRC="FIGDIR/small/605499v1_ufig1.gif" ALT="Figure 1"> View larger version (33K): org.highwire.dtl.DTLVardef@1bb166forg.highwire.dtl.DTLVardef@1bde2b6org.highwire.dtl.DTLVardef@16f1b9eorg.highwire.dtl.DTLVardef@17e4ff1_HPS_FORMAT_FIGEXP M_FIG C_FIG

microbiology↗

bioRxiv2024

AlphaFold Model Quality Self-Assessment Improvement Via Deep Graph Learning

In the past several years significant advances have been made in the field of deep-learning based computational modeling of proteins, with DeepMinds AlphaFold2 being among the most prominent. Alongside the atom coordinates, these computationally modeled protein structures typically contain self-confidence metrics that can be used to gauge the relative modeling quality of individual residues, or the protein as a whole. Unfortunately, these scores are not always accurate, and may sometimes annotate poorly modeled regions of the protein as high confidence. Here, we introduce EQA-Fold to address this problem. EQA-Fold overhauls the LDDT prediction head of AlphaFold to provide more accurate self-confidence scores. We show that EQA-Fold is able to provide more accurate self-confidence scores than the standard AlphaFold architectures, as well as recent Model Quality Assessment protocols.

bioinformatics↗

bioRxiv2024

Predicting transcriptional activation domain function using Graph Neural Networks

Analysis of factors that lead to the functionality of transcriptional activation domains remains a crucial and yet challenging task owing to the significant diversity in their sequences and their intrinsically disordered nature. Almost all existing methods that have aimed to predict activation domains have involved traditional machine learning approaches, such as logistic regression, that are unable to capture complex patterns in data or plain convolutional neural networks and have been limited in exploration of structural features. However, there is a tremendous potential in the inspection of the structural properties of activation domains, and an opportunity to investigate complex relationships between features of residues in the sequence. To address these, we have utilized the power of graph neural networks which can represent structural data in the form of nodes and edges, allowing nodes to exchange information among themselves. We have experimented with two kinds of graph formulations, one involving residues as nodes and the other assigning atoms to be the nodes. A logistic regression model was also developed to analyze feature importance. For all the models, several feature combinations were experimented with. The residue-level GNN model with amino acid type, residue position, acidic/basic/aromatic property and secondary structure feature combination gave the best performing model with accuracy, F1 score and AUROC of 97.9%, 71% and 97.1% respectively which outperformed other existing methods in the literature when applied on the dataset we used. Among the other structure-based features that were analyzed, the amphipathic property of helices also proved to be an important feature for classification. Logistic regression results showed that the most dominant feature that makes a sequence functional is the frequency of different types of amino acids in the sequence. Our results consistent have shown that functional sequences have more acidic and aromatic residues whereas basic residues are seen more in non-functional sequences.

bioinformatics↗