Virtual-SAGE: a new approach to EST data analysis.

We present a computer program, termed V-SAGE (Virtual-SAGE), designed to facilitate the analysis of gene expression profiles by combining elements of SAGE (Serial Analysis of Gene Expression) with high-throughput EST analysis. The program re-iteratively correlates sequence tags adjacent to poly(A) tail sequence strings with a second or several tags located within the cDNA adjacent to the recognition sequences of frequently-cutting endonucleases. By recording tags and their distance, the program generates an expression profile from large numbers of sequences, groups sequences according to tags, and identifies alternatively spliced transcripts as well as transcripts that are characterized by 3'-UTR sequences of different length. We discuss the application of V-SAGE to a collection of corn root segment transcripts.