We have surveyed more than 374,000 SAGE tags representing 66,502 unique transcripts from nine different lung tissue libraries. These libraries included two of each from normal small or large airway epithelial cell cultures, two of each from primary squamous cell lung cancers and adenocarcinomas of the lung, and a 9th library derived from the A549 lung adenocarcinoma cell line. We assigned the identified SAGE tags to their chromosomal positions using the newly completed human genome sequence and developed a genome-wide gene expression profile representing the relative expression pattern of each transcript among the nine lung tissues. Of the 66,502 unique transcripts, 23,056 transcripts appeared more than once among all nine libraries, 18,652 (81%) of those matched to UniGene clusters while 85 tags (0.4%) matched to mitochondrial DNA/ribosomal RNA sequences. Excluding the 4,938 tags (21.4%) that matched to more than one cluster, 13,714 tags (59.5%) were linked to unique or positionally related UniGene clusters and included in the transcriptome map. The resulting transcriptome map revealed clusters of highly differentially expressed genes at specific chromosomal regions of the genome and suggested that the generation of transcriptome map using SAGE may provide a potentially powerful tool to help reveal candidate regions of chromosomal changes in lung cancer.
