Table 1.

Characteristics of Gene Indexing Databases and Classification Methods

Feature		Databases and clustering methods
Feature		TGI	UniGene	STACK	GeneNest	Mouse cDNA project	Rice cDNA project (our method)
Clustering program	`MegaBlast` and`CAP3`	`MegaBlast`	`d2_cluster`	`BLAST`	`BLAST`	`BLAST`
Representative clones	No	Yes	No	Yes	Yes	Yes
Consensus sequences	Yes	No	Yes	Yes	no	No
Alignments	Yes	No	Yes	Yes	Yes	No
Alternative splices	Different clusters	Same clusters	Different clusters	Same clusters	Same clusters	Same and different clusters
Redundancy of groups	High	Low	High	Low	Low	Low
Visualization tool	Yes	No	Yes	Yes	Yes	No
Other information	Coding potential	Annotation tissue	Tissue	Sequence quality	Tissue	Tissue

↵ http://www.tiogr.org/tdb/tgi.shtml. Quackenbush et al. 2001.
↵ http://www.ncbi.nlm.nih.gov/UniGene/index.html. Schuler et al. 1997.
↵ http://www.sanbi.ac.za/Dbases.html. Christoffels et al. 2001. Burke et al. 1998.
↵ http://genenest.molgen.mpg.de/. Haas et al. 2000.
↵ Konno et al. 2001.
↵ http://www.ncbi.nlm.nih.gov/blast/
↵ Huang et al. 1999.
↵ Burke et al. 1999. Four gene indexing databases and two classification methods used for our full-length cDNA projects are compared based on eight distinct points of view.

A Computer-Based Method of Selecting Clones for a Full-Length cDNA Project: Simultaneous Collection of Negligibly Redundant and Variant cDNAs