The UK Crop Plant Bioinformatics Network (UK CropNet)

UK CropNet currently provides a range of databases (and database-mining tools) to the plant community that are all freely accessible through our website (http://ukcrop.net/). Recent upgrades have meant that we can now expand the range of available facilities (e.g. addition of new databases) whilst also strengthening and improving access to existing services (e.g. providing a BLAST search facility against sequences in our databases). This article will briefly outline these and other new developments in our service.


Faster machine, faster access
Several practical improvements are immediately available due to the increased memory, processing power, and disk space on our new UK CropNet server.For example, we can now handle multiple database connections more ef®ciently, providing an increase in the apparent `speed' to the end user.This is most likely to be noticed when many users are connecting simultaneously.
Our increased disk space now allows UK Crop-Net to mirror many more databases from other sources (principally from Cornell University).The eighteen existing ACEDB plant-related databases (®ve from CropNet, thirteen mirrored from Cornell) are joined by an additional thirteen databases.These include many new databases for model organisms (e.g.MaizeDB) and other globally important plant species (e.g.CassavaDB).We now also have several databases that are quite atypical of most ACEDB databases, i.e. they don't contain any sequences or maps.Such a database is EcoSys, a database that contains details on the natural environmental conditions (temperature, soil pH, precipitation, etc.) for nearly 900 different taxa, arranged by family and genus.
Database searching is now much faster which has made it possible to develop a prototype multipledatabase-search facility (see: http://ukcrop.net/perl/ace/mgrep/ or http://ukcrop.net/db.html).This facility will allow a user to perform a text-based query search against any (or all) of our databases (Figure 1), i.e. a user can search for the name of a gene of interest across all of the databases at our site.

Improvement to web-based database interfaces
Accessing an ACEDB-style database via the web can sometimes prove unsatisfactory when compared to using the database on a local workstation.This is because the functionality of a complex database program such as ACEDB is compromised by the restrictions of what you can and cannot do within the con®nes of an HTML web page.To address this, we have introduced a number of new mod-i®cations and tweaks to the standard package of ACEDB web interfaces (collectively known as `AceBrowser').At a very simple level, some of these `tweaks' involve nothing more than the simple rewording of much of the default AceBrowser text to something that is (hopefully) more lucid and intuitive to the end user (particularly when English might not be their ®rst language).At a more advanced level, we have re®ned many of the AceBrowser interfaces and, where necessary, developed and added completely new interfaces.
For instance, several of the CropNet databases (AGR, BrassicaDB, MilletGenes) are now linked to a Java based map display (Pairwise Comparative Map, PCM) that offers vastly improved functionality over the standard ACEDB web interface for viewing maps.Users can now quickly resize, zoom, and invert maps and compare multiple maps in the same, interactive, display (where multiple maps exist).
Other modi®cations to the ACEDB web interfaces include increasing the amount of information available within graphic displays of sequences (useful when viewing sequences which have a lot of associated homology information from `BLAST hits') and providing access to the ACEDB `Table-Maker' facility for data organisation through the standard web interface.It has also been possible to con®gure the database displays to allow for interdatabase linking, e.g. a BLASTX homology from a nucleotide sequence in BrassicaDB can now be linked to the corresponding protein sequence in AGR.
One of the major modi®cations to the AceBrowser interface has been the addition of a BLAST search facility.When browsing any (sequencecontaining) database, users can now access a BLAST page (Figure 2), which provides a search facility for any of these databases.BLAST databases are updated each night to re¯ect any new sequence additions.For Arabidopsis, it has also been possible to create multiple BLAST databases to re¯ect the fact that the Arabidopsis genome sequence is nearly complete.Users can thus search against all Arabidopsis DNA (y250,000 sequences) or only against those sequences that comprise the Arabidopsis Genome Initiative (AGI) genome sequence.Even more speci®cally, users can search against AGI sequences that have been mapped to each of the ®ve chromosomes.Currently, a choice of BLASTN, BLASTP, and BLASTX is available depending on whether protein or nucleotide sequences exist for a given database.
Early in 2001 we will be introducing a generic Java based access tool developed within UK CropNet as an alternative method for browsing our databases.This tool is `GFace' and provides an opportunity to utilise realtime manipulations of data via the web that eliminates many of the refresh issues and some of the lack of functionality that has been inevitable through simple HTML based browser access.GFace has been speci®cally designed to incorporate all of UK CropNet's Java tools (e.g.PCM mentioned above) and is also designed to integrate with tools (under development) for the cross-interrogation of ACEDB style databases with other plant database formats such as Oracle.
On a ®nal note, we are constantly striving to make our service as accessible and powerful for the widest possible scienti®c audience.This ranges from making our databases easier to use and the data more conveniently visualised, to incorporating `onthe-¯y' translations for several European languages via `babel-®sh' from alta-vista.As part of this endeavour, we welcome any and all comments and criticisms (preferably in English) that may help to improve and develop UK CropNet in the future.

Figure 1 .
Figure 1.The multiple database search tool

Figure 2 .
Figure 2. The BLAST search page for AGR