The bioinformatic catalyst in the kallikrein family.

Broadly speaking, bioinformatics is simply the application of computers to solve biological problems. In this context, bioinformatics has been around for decades. Although more specific definitions for bioinformatics will vary, the National Center for Biotechnology Information proposed that bioinformatics represents the field of science wherein biology, computer science and information technology merge into a single discipline. An overview of the more commonly used bioinformatic methods and their applications is described elsewhere [1]. Bioin-formatics only really came to the fore after the initiation of the human genome project in 1988. This was a driving force in the development of databases to store and compare the huge amount of sequence data that was being generated throughout the 90s. Hand in hand with this burgeoning wealth of data were the essential advances in bioinformatic tools. Together these facilitated the comparison and search analyses required to add predictive clinical value to the data. Nowadays, any traditional approach to analyse this type of data would be like trying to cut down a forest with a hacksaw. Bioinformatics, data mining or the more fashionable term, in silico analyses, represents the only way to make sense of the huge volume of information available. Consequently, articles based solely on bioinformatic or in silico type of analyses are becoming more frequent. Despite the apparent 'plug-and-play' simplicity of the in silico approach, it still requires the same attention to detail as say choosing an appropriate statistical analysis or a 'wet-bench' method. Contamination of database sequences, redundancy, frame-shifts and errors in genome annotations are just a few of the problems previously encountered. Some important issues related to these studies are addressed in a review by Altschul et al. [2], but the literature is replete with tips and suggestions for fine tuning data searches for speed and accuracy. In this issue of Tumor Biology, Yousef et al. [3] have performed an impressive bioinformatic analysis on the newly characterized human kallikrein gene hKLK6. The same group has contributed significantly to the kallikrein cause in recent years with a number of publications using traditional 'wet' biology [4–9] and they have been able to substantiate many of their assertions in the current in sili-co study [3]. This group, led by Prof. E.P. Diamandis in Toronto, are no strangers to the bioinformatic approach [10–12]. Using 11 databases, they have examined the gene structure of KLK6, extending the clone sequence by comparing overlapping clones (they identified six …