“In modern molecular biology, research scientists often want to investigate the action of large numbers of the genes in a given species in a systematic and high-throughput manner (so-called functional genomics), and this paper describes the creation of a novel resource for Xenopus tropicalis researchers which allows them to do that in a very convenient way.”

I think it is highly cited because it represents a major step in the development of functional genomics resources for the rapidly emerging model organism Xenopus tropicalis. Because of the work described in the paper, researchers using this species now have access to both a highly diverse physical reagent set, immediately suitable for large-scale expression screens, and a database with which to verify the sequence, identity, and other characteristics of the reagents. There is also a nice match between the focus of most Xenopus research on early development, and the particular cDNA libraries on which the work was based. The recent publication of the X. tropicalis genome assembly by the Joint Genome Institute in Walnut Creek, California, gives an indication of the increasing interest in the use of this amphibian species to help unravel the basic mechanisms of vertebrate embryonic development.

There are two new methodologies presented in the paper: a scaleable EST (expressed sequence tag) clustering algorithm which avoids some of the pitfalls of current methods, and an approach to selecting putative full-length clones which has a high success rate and which may occasionally find genes which have so far not been found in other species. The clustering algorithm is uniquely able to produce assembled contigs from very large numbers of ESTs (up to at least one million) while avoiding the common problems of chimeric mis-assembly, and (computer) memory exhaustion when very large numbers of similar sequences are being analyzed. The EST clustering is already proving useful to other researchers, having been adopted by a group of scientists in Denmark to analyze a very large collection of pig ESTs, and the set of putative full-length clones has been fed into the Sanger Institute's full-insert cDNA sequencing program and is now providing accurate full-length sequence information and identification for large numbers of Xenopus genes. Some of the ideas introduced here may also be useful to other bioinformaticians working with sequence assembly and clone-picking algorithms.

Xenopus tropicalis (a small African clawed frog) is an ideal model organism with which to investigate fundamental mechanisms of early embryonic development in vertebrates. In modern molecular biology, research scientists often want to investigate the action of large numbers of the genes in a given species in a systematic and high-throughput manner (so-called functional genomics), and this paper describes the creation of a novel resource for Xenopus tropicalis researchers which allows them to do that in a very convenient way. It also describes a new method of sorting the hundreds of thousands of short sequences (ESTs) that are used to build up the picture of which genes are functioning during these stages of development of the animal—from egg to tadpole.

I work in the field of bioinformatics at the The Gurdon Institute in Cambridge, and a large part of the role is to use my computational expertise to complement the experimental skills of my colleagues in collaborative projects. This is particularly true where the project requires the development of large scale computational analyses of experimental data. In this case my primary collaborators in the Institute had a very good idea of what they wanted to get out of the enormous quantities of EST data that had been generated, but no real way to make it happen. My arrival at the Institute happily coincided with the completion of the first phase of EST sequencing and I was rapidly able to organize the data and begin extracting the information they were looking for.