The KEGG resource is a comprehensive bioinformatics database that integrates genomic, chemical, and network information to help decipher the genome. It includes three main databases: the gene universe (GENES, SSDB, KO), the chemical universe (COMPOUND, GLYCAN, REACTION), and the protein network (PATHWAY). These databases represent three graph objects: the gene universe, the chemical universe, and the protein network. The gene universe represents ortholog/paralog relations, operon information, and gene relationships. The chemical universe represents chemical reactions and structural/functional relations among metabolites and other biochemical compounds. The protein network represents known molecular interaction networks in various cellular processes. KEGG is an integrated resource that provides a self-sufficient system for linking genomes to life at the cellular level, containing a complete set of building blocks (genes and molecules) and wiring diagrams (interaction networks) for cellular functions. It is used for analysis of various types of molecular biological data to obtain clues to higher-order functions. The KEGG resource has been made more accessible to automated analysis, with the XML representation of pathway diagrams and the KEGG API (SOAP interface) allowing users to write programs for analyzing microarray gene expression data or annotating newly sequenced genomes. New database components include KO for ortholog grouping and hierarchical classification of genes and GLYCAN for carbohydrate structures. The KEGG databases are described in detail, including the COMPOUND, GLYCAN, REACTION, and ENZYME databases. The PATHWAY database contains manually drawn diagrams called KEGG reference pathway diagrams, each corresponding to a known network of functional significance. The KEGG Markup Language (KGML) is used for computational reconstruction and manipulation of KEGG pathways. Access to KEGG is available through the GenomeNet website and via the KEGG API and anonymous FTP. The KEGG resource is supported by grants from various Japanese organizations.The KEGG resource is a comprehensive bioinformatics database that integrates genomic, chemical, and network information to help decipher the genome. It includes three main databases: the gene universe (GENES, SSDB, KO), the chemical universe (COMPOUND, GLYCAN, REACTION), and the protein network (PATHWAY). These databases represent three graph objects: the gene universe, the chemical universe, and the protein network. The gene universe represents ortholog/paralog relations, operon information, and gene relationships. The chemical universe represents chemical reactions and structural/functional relations among metabolites and other biochemical compounds. The protein network represents known molecular interaction networks in various cellular processes. KEGG is an integrated resource that provides a self-sufficient system for linking genomes to life at the cellular level, containing a complete set of building blocks (genes and molecules) and wiring diagrams (interaction networks) for cellular functions. It is used for analysis of various types of molecular biological data to obtain clues to higher-order functions. The KEGG resource has been made more accessible to automated analysis, with the XML representation of pathway diagrams and the KEGG API (SOAP interface) allowing users to write programs for analyzing microarray gene expression data or annotating newly sequenced genomes. New database components include KO for ortholog grouping and hierarchical classification of genes and GLYCAN for carbohydrate structures. The KEGG databases are described in detail, including the COMPOUND, GLYCAN, REACTION, and ENZYME databases. The PATHWAY database contains manually drawn diagrams called KEGG reference pathway diagrams, each corresponding to a known network of functional significance. The KEGG Markup Language (KGML) is used for computational reconstruction and manipulation of KEGG pathways. Access to KEGG is available through the GenomeNet website and via the KEGG API and anonymous FTP. The KEGG resource is supported by grants from various Japanese organizations.