Systematic analysis of an organism’s transcript repertoire plays a key role in molecular genetics. The complexity of this task in higher organisms such as mammals is not only owing to the large number of expected genes (70,000–100,000) in humans (1) but also because gene coding regions are dispersed in the genome. In addition, only a subset of all transcripts is found in a given cell type, typically in the range of 10,000, and the presence of certain transcripts is subject to the developmental stage and/or the cell’s response to stimuli. The idea of a systematic approach to sequence large numbers of DNA (reverse transcribed from mRNA) clones as part of the Human Genome Project was conceived in 1990 at a time when significant advances in the field of DNA sequencing technology were in place but fewer than 2000 unique human gene sequences were available in the public databases. In 1991, Adams et al. (2) reported the generation of 174,172 human partial cDNA sequences and introduced the term expressed sequence tags (ESTs). Since then, other major sequencing efforts (3) have brought the number of human ESTs in databases (4) to more than 1,400,000.