GeneNest: automated generation and visualization of gene indices.

Expressed sequence tags (ESTs), introduced by Adams et al. in 1991, are a rapidly growing resource for analyzing genes. Although ESTs may be of low sequence quality they are extremely useful for detecting new genes, determining the genomic structure of a gene (exon−intron boundaries, alternative splicing) or for expression studies. Since EST sequence information is highly redundant a single gene may be covered by thousands of ESTs each representing different parts of that gene. In order to simplify the analysis of specific genes several efforts have been made to cluster sequences belonging to the same gene resulting in so− called gene indices. Some commonly used gene indices are Unigene at NCBI (National Center for Biotechnology Information), the TIGR (The Institute for Genomic Research) gene indices and STACK at SANBI (South African National Bioinformatics Institute). In particular Unigene and the TIGR gene indices differ mainly in the clustering strategy used and in the presentation of cluster related information. Clusters of the TIGR gene indices are summarized by a database of consensus sequences each reflecting a single transcript. Additionally, the relativ order of sequences within a cluster is roughly sketched. In contrast, sequences in Unigene are clustered less stringently such that alternative splice variants fall into the same cluster. Sequences derived from the same clone also may be clustered only based on their annotation. Unigene focuses on an