Searches against STACK can yield extended consensus sequences made from constituent ESTs, which in turn yield longer queries for subsequent work. Links and consensus sequences make the process of working with ESTs more simplified, and in some cases, purified entries in the form of SANIGENE can be extracted and used. Searches using the SANBI search engine against STACK 2.0 are linked directly to an ENTREZ linked entry retrieval engine, thus improving the access to other databases such as UniGene and dbEST. A Virtual Transcribed Sequences set of gene-fragments that have matched STACK entries from available human genome sequences will shortly be linked to the STACK-SEARCH engine at SANBI. Drosphila ESTs that have found matches with STACK entries can be found at a server at TigemNet (see end of article).

STACK alignments are extremely useful for cSNP discovery, and for provision of scaffolds for further assembly via use of STACK consensi. Occasionally, errors have occurred in manufacture, and a collection of non-included ESTs is distributed with the database. A 2.1 release will contain the adjusted entries.

STACK 2.0 is now available on our FTP site (academic registration for download: http://ziggy.sanbi.ac.za/stack/stackrequest.htm) and can be searched on our BLAST engine.

STACK 2.0 has been generated from genbank103 release. It comprises 16 tissue sets. This differs from the previous release as our previous synovial membrane set has been incorporated into the connective tissue set.

Genbank103 release was clustered on an SGI origin2000 as well as a Maspar. No claims are made as to the accuracy of the following statistics. We welcome comments and corrections.