The Arabidopsis Information Resource (TAIR) is the model organism database for Arabidopsis thaliana, a widely used plant model. TAIR provides genomic and genetic data, including genes, clones, ecotypes, markers, expression data, SNPs, mutant alleles, phenotypes, proteins, and sequences. It also includes publications, protocols, researchers, and labs. TAIR integrates data from the Arabidopsis Biological Resource Center (ABRC), allowing users to discover and access seed and DNA stocks. TAIR has seen increased usage, with over 110,000 visits in a single month. TAIR data sources include manual curation of research literature, computational pipelines for gene structure and function annotation, and data imports from GenBank and ABRC. Manual curation focuses on high-impact journals, while computational methods generate gene structure and functional annotations. TAIR has added the GBrowse genome viewer, redesigned its interface, and launched several web services. The latest release, TAIR7, includes 27,029 protein-coding genes, 3,889 pseudogenes or transposable elements, and 1,123 ncRNAs, with 37,019 gene models. TAIR's annotation pipeline combines manual and automated methods. Automated updates are reviewed manually, with some rejected due to errors. TAIR has corrected 1,098 gene models for TAIR7, improving gene structure accuracy. The TAIR7 release includes 10,792 updated gene structures, with 41 gene splits and 34 gene merges. The average 3' UTR length decreased, while the 5' UTR increased. TAIR has expanded its annotation to include small proteins and non-coding transcripts, incorporating 467 cysteine-rich peptides and 213 genes as 'other_RNA'. It also includes alternative spliced genes, with 1,002 new splice variants. Functional annotations use GO terms and computational methods, with 60% of genes annotated to a GO molecular function, 50% to a biological process, and 49% to a cellular component. Future TAIR releases will refine gene structures, improve transposable element and pseudogene annotations, and incorporate more splice variants. TAIR plans to release new genome versions at least once a year, with TAIR8 expected in early 2008. TAIR is supported by the National Science Foundation.The Arabidopsis Information Resource (TAIR) is the model organism database for Arabidopsis thaliana, a widely used plant model. TAIR provides genomic and genetic data, including genes, clones, ecotypes, markers, expression data, SNPs, mutant alleles, phenotypes, proteins, and sequences. It also includes publications, protocols, researchers, and labs. TAIR integrates data from the Arabidopsis Biological Resource Center (ABRC), allowing users to discover and access seed and DNA stocks. TAIR has seen increased usage, with over 110,000 visits in a single month. TAIR data sources include manual curation of research literature, computational pipelines for gene structure and function annotation, and data imports from GenBank and ABRC. Manual curation focuses on high-impact journals, while computational methods generate gene structure and functional annotations. TAIR has added the GBrowse genome viewer, redesigned its interface, and launched several web services. The latest release, TAIR7, includes 27,029 protein-coding genes, 3,889 pseudogenes or transposable elements, and 1,123 ncRNAs, with 37,019 gene models. TAIR's annotation pipeline combines manual and automated methods. Automated updates are reviewed manually, with some rejected due to errors. TAIR has corrected 1,098 gene models for TAIR7, improving gene structure accuracy. The TAIR7 release includes 10,792 updated gene structures, with 41 gene splits and 34 gene merges. The average 3' UTR length decreased, while the 5' UTR increased. TAIR has expanded its annotation to include small proteins and non-coding transcripts, incorporating 467 cysteine-rich peptides and 213 genes as 'other_RNA'. It also includes alternative spliced genes, with 1,002 new splice variants. Functional annotations use GO terms and computational methods, with 60% of genes annotated to a GO molecular function, 50% to a biological process, and 49% to a cellular component. Future TAIR releases will refine gene structures, improve transposable element and pseudogene annotations, and incorporate more splice variants. TAIR plans to release new genome versions at least once a year, with TAIR8 expected in early 2008. TAIR is supported by the National Science Foundation.