The cell cycle database is accessible at (http://www.itb.cnr.it/cellcycle).The web interface is made up of a set of HTML pages dynamicallygenerated from PHP scripts, in order to retrieve informationabout genes and proteins related to the cell cycle process ina specific report created for each gene and protein. Moreover,a search related to the cell cycle models stored in the databaseis possible: users can retrieve the list of the mathematicalmodels, choose one of them and visualize the related informationon the web pages.
Gene and protein report
Users browsing the website searching for annotations of thecomponents of the cell cycle are redirected to specific reports.
Besides the common query possibilities (by gene/protein name,by keywords and by IDs of most common biological resources),users can query the database using the BLAST algorithm (8),which is particularly useful in order to discover similaritiesamong unknown cell cycle putative genes and the database content.
The gene report lists all the information related to each genethat is stored in the database, starting from the basic genedescription, its sequence and its corresponding protein, butit also includes more specific information, such as the listof the SNP characterizing that gene, or the list of cDNA andisoform. Furthermore, in the gene report, particular attentionis given to the information related to the promoter regionsand to the transcription factors specific for each yeast andhuman gene, in order to facilitate research on cell cycle generegulation. We also provide links to experimental data on geneexpression taken from the gene expression omnibus (GEO) repositoryin order to present as much supplementary information as possibleconcerning the cell cycle genes. Since the regulation of cyclin-dependentkinases (CDK) characterizes the most crucial events of the cellcycle (9), we supply additional information about kinase genesby using the link to the KinWeb database (10).
As far as the protein report is concerned, particular attentionis given to the network of protein–protein interactionsinvolved in the cell cycle. The database contains protein–proteininteractions taken from several resources making the informationon the cell cycle interaction network as complete as possible.In the protein report, the graphical visualization of the domainsfrom the InterPro database is provided. Users can also directlyvisualize the protein structure and the related Connolly surface(11) according to PDB data, using the Java 3D applet. Moreover,for each protein we provide information on the models in whichit is involved, a list of the published models is availabledirectly in the protein report with a direct link to the specificmodel report discussed in the following section.
More information both for genomics and proteomics integratedand linked resources is available in the help pages of the cellcycle DB website.
Using the web interface it is possible to retrieve model-relatedinformation and a pipeline has been implemented to deal withthe mathematical part of the models, in order to solve the ordinarydifferential equations systems that describe the biologicalprocesses. Using this system it is possible both to visualizethe mathematical description of the model and to run simulationsvarying initial conditions of state variables and parameters.
Each model is presented in a report structured in three sections:the publication data, the SBML data structure and the numericalsimulation part. The first section contains the detailed publicationdata, the diagram of the model, the related XML file and thelist of all the proteins involved in the model that are linkedto the related cell cycle database protein report.
In the SBML data structure section, users can explore the SBMLcomponents of the selected model including its mathematicalexpressions. Mathematical formulas within the SBML models areexpressed using the Mathematical Markup Language (MathML orMML) (12). To view the expressions on the web, our resourcerelies on a XHTML + MathML page. This technology allows thegeneration of high-quality documents in which mathematical expressionsare treated as text in a HTML web page: it is possible to usethe browser functions to find strings inside mathematical expressionsand to change their size, operations that are usually not possibleusing images to represents formulas (Figure 2a).
The simulation section allows users to simulate a model usingthe software XPPAUT (13) and to plot results on the fly in orderto capture the dynamical properties of the biological process.In order to enable the simulation, this section lists the modelspecies (state variables, typically protein concentrations),its parameters (such as kinetic constants), its algebraic rulesand XPPAUT internal options, using default values. Users canchange the initial values in order to analyse the differentnatures of the system dynamics, performing sensitivity and bifurcationsanalysis. Once the computation is completed, users can downloadXPPAUT input and output files and plot results, as shown inFigure 2b. The web interface allows users to plot both timecourses and phase diagrams for each model after the simulationstep. Results are shown with images exported by GNUPLOT (14),the popular portable command-line function plotting software.