Biology subjects

Kanehisa, M.

Publications and source records attributed to Kanehisa, M..

4 recordsLinked to original sources

bioRxiv2026

A hierarchical orthology framework reveals viral carbohydrate-active genes across the global virosphere

Carbohydrate-active enzymes (CAZymes) shape virus-host interactions by modifying virion structures, host surfaces and extracellular glycans. However, the diversity and evolutionary origins of viral carbohydrate-active enzymes remain poorly understood, partly due to limited viral protein annotations. To address this, we present VirGenes, a database of viral orthologous groups constructed from the KEGG viral gene dataset. VirGenes uses a hierarchical framework that integrates sequence similarity, remote homology, and structural similarity to support evolutionary and functional analyses of viral proteins. By screening the sequence space of VirGenes, we identified 558 CAZyme-associated gene clusters spanning 102 CAZyme families, revealing particularly enriched repertoires in dsDNA viral lineages. Two bacteriophage families, Kleczkowskaviridae and Pootjesviridae, encoded more than 10 CAZymes per genome, followed by Mimiviridae, a representative family of eukaryotic giant viruses. Phylogenetic analyses systematically revealed divergent evolutionary histories of viral carbohydrate-active genes, including frequent horizontal transfer of endolysin genes from bacteria, which likely represents a viral strategy in the ongoing evolutionary arms race with their cellular hosts. Within the structural space of VirGenes, a large number of viral genes were found to contain CAZyme-like folds despite more than 85% of them lacking detectable sequence similarity to annotated CAZyme sequences. Notably, numerous hypothetical sequences from giant viruses exhibited glycoside hydrolase-like five-bladed {beta}-propeller folds. Overall, by integrating sequence, structural and functional evidence, we show that viral carbohydrate-active systems exemplify how distributed innovations, constrained by ancient folds, collectively build the functional complexity of the global virosphere. VirGenes is publicly accessible at https://www.genome.jp/vogdb/.

bioinformatics↗

bioRxiv2026

DeepKOALA: A Fast and Accurate Deep Learning Framework for KEGG Orthology Assignment

The KEGG Orthology (KO) system links DNA and protein sequences to biological functions and pathways, providing a curated, fundamental and consistent annotation framework across all domains of life. While accurate, traditional sequence alignment-based annotation methods are computationally expensive, which severely limits their application in large-scale datasets. To address this challenge, we introduce DeepKOALA, a deep learning approach based on Gated Recurrent Units (GRU), which frames KO annotation as an open-set recognition task. This design reduces false positives arising from out-of-scope sequences and, together with a lightweight GRU backbone, enables high-throughput annotation. The GRU-based model was benchmarked against four other deep learning architectures and showed the best balance between speed and accuracy. We further performed a cross-species evaluation of DeepKOALA in comparison with existing KO annotation tools, where it achieved an F1-score of 0.8653 and 37.5-fold acceleration compared with BlastKOALA. We also provide a specialized fragment model for handling incomplete sequences and an optional Multi-domain mode. Together, these features make DeepKOALA a scalable, efficient, and accurate solution for high-throughput function annotation. O_FIG O_LINKSMALLFIG WIDTH=200 HEIGHT=61 SRC="FIGDIR/small/698072v1_ufig1.gif" ALT="Figure 1"> View larger version (15K): org.highwire.dtl.DTLVardef@8c3b53org.highwire.dtl.DTLVardef@8acfa8org.highwire.dtl.DTLVardef@147398borg.highwire.dtl.DTLVardef@112d762_HPS_FORMAT_FIGEXP M_FIG C_FIG

bioinformatics↗

bioRxiv2019

KofamKOALA: KEGG ortholog assignment based on profile HMM and adaptive score threshold

SummaryKofamKOALA is a web server to assign KEGG Orthologs (KOs) to protein sequences by homology search against a database of profile hidden Markov models (KOfam) with pre-computed adaptive score thresholds. KofamKOALA is faster than existing KO assignment tools with its accuracy being comparable to the best performing tools. Function annotation by KofamKOALA helps linking genes to KEGG resources such as the KEGG pathway maps and facilitates molecular network reconstruction.\n\nAvailabilityKofamKOALA, KofamScan, and KOfam are freely available from https://www.genome.jp/tools/kofamkoala/\n\nContactogata@kuicr.kyoto-u.ac.jp

bioinformatics↗

bioRxiv2017

TOWARDS A GLOBAL SUPPORT OF CORE DATA RESOURCES FOR THE LIFE SCIENCES

On November 18-19, 2016, the Human Frontier Science Program Organization (HFSPO) hosted a meeting of senior managers of key data resources and leaders of several major funding organizations to discuss the challenges associated with sustaining biological and biomedical (i.e., life sciences) data resources and associated infrastructure. A strong consensus emerged from the group that core data resources for the life sciences should be supported through a coordinated international effort(s) that better ensure long-term sustainability and that appropriately align funding with scientific impact. Ideally, funding for such data resources should allow for access at no charge, as is presently the usual (and preferred) mechanism. Below, the rationale for this vision is described, and some important considerations for developing a new international funding model to support core data resources for the life sciences are presented.

scientific communication and education↗