Cyclonet database consists of several sections:

• BioUML data - tables used by BioUML workbench for diagrams and biological components description

• Microarray - microarray data related with cell cycle and cancer

• Publications - references and full articles concerning cell cycle regulation, cancer and drugs.

• Resources - list of miscellaneous useful resources.

• Chemoinformatics - issues related with chemoinformatics. This section will reuse some BioUML data.

• Organisational issues - teams, people, blogs and other

Cyclonet data is stored on the server side in relational databases. MySQL database (http://www.mysql.com) is used for this purpose. There are two ways how user can get access to the database content:

1) Cyclonet module – provides integration of Cyclonet database into BioUML workbench. The module allows users to query and edit the database content, view and edit diagrams, analyze and simulate the described systems behavior using MATLAB or own BioUML simulation engine.

2) Using conventional web browser user can query and edit the database content as well as view diagrams. BeanExplorer Enterprise Edition (EE) technology is used for this purpose. BeanExplorer generates complete user interface of web application directly from the database metadata without the need for manual programming of database and interface objects.

See the MS Word document for more details: Cyclonet.doc

Figure above shows current Cyclonet database strucrure as ER diagram. Boxes corresponds to tables and shows corresponding table fields, edges shows relationships/dependencies between tables.

BioUML data

BioUML SQL module is used to store BioUML data in relational database. 

This module is also can be used by other databases. Now we are also developing Biopath (codename) other database using BioUML technology.

 This database will store information about processes in organism due to excess of heavy metals, development of conceptual model of pathogenesis chronic noninfectious lung diseases and optimization of their treatment. While BioUML section in Cyclonet database is empty, you can take a look on diagrams and component descriptions in biopath database:  http://biopath.biouml.org 

Generally BioUML data consists of

• entities - different types of biological components and concepts that can be used as graph nodes in a diagram. Currently following tables are used:

  • concepts - this term is used to describe biological function/activity, process or state.
  • cells - description of cells, and cell lines
  • compartments - description of cellular and organism compartments
  • substances - any substance that is not biopolymer
  • genes - gene
  • rna - RNA
  • proteins - protein or protein complex
     
• references - any BioUML entity has synonims, publicationReferences and database references attributes. While it is 1..n or n..m relationship several auxiliary tables are used:
  • publicationReferences
  • synonyms
  • databaseReferences
     
• relationships. There are two types of relationships between entities in BioUML
  • semantic relationship - binary relation to indicate that two entities are related (is-a, part-of, etc)
  • reaction - special node type whose edges linked with reaction reactants, products and modifiers.
     

• substanceAction - we can try to extend semantic relationship for this purpose. This will allow us to embed chemoinformatics data into BioUML workbench. For example we can it
   

•  different dictionaries and utility tables:
  • constants - here we plan to store information about reaction rate constants, compartment volumes, concentrations of different substances in normal conditions and pathology, etc.
  • units - measurement units, like second, liter, etc.
  • species - list of used species
  • ...

BioUML workbench - new version (v. 0.7.2) of BioUML workbench is required to work with Cyclonet database through the Internet.

Microarray data

this section of database is used to store microarray data related with cell cycle and cancer and results of their analyses.

Currently this section includes:

• series - list of experiments (series) related with cell cycle and cancer

• seriesReferences - set of articles stored in publications table and linked with corresponding experiments

• resources - contains resources related to microarray experiments.
   

• mGenes - the purpose of this table is to provide common gene identifier. The table will contain complete list of human genes. This information will be extracted from LocusLink, if microarray gene chips will contain probes missing in LocusLink and probe corresponds more or less known gene, we will use information from Unigene or GenBank.
  Note: The database also contains gene tables and is used by a BioUML module. The mGene (microarray gene) table will be linked with the genes table.

• there will be a set of tables that will link our mGene identifiers with LocusLink, Unigene, etc.

  • mGene2unigene
  • mGene2locusLink
  • mGene2affymetrix
  • .. additinal tables can be added if it will be necessary
     

•  affymetrix - data about affymetrix probes. Currently this table contains probesID  (about 300 000) for 17 human platforms. Table below summarise statistics about Affymetrix platforms and number of probesID linked with other resources: LocusLink, UniGene, RefSeq and ensembl.
   

• groups - groups and related tables are used to describe results of analysis of microarray data. For example, one article can provide a list of genes with cyclic expression during cell cycle -  this will be a group. Additionally these genes also can be grouped by functions, in this case we will add a set of groups - one group for each function. Other example when authors identify group of overexpressed and underexpressed genes for some conditions, these results will be stored as two groups.  

• gene_groups - will contain a list of genes for the specified group. Additionally several values can be associated with gene, the value meaning is described in groups table. In first case it can be periodicity score, in second example it can be level of gene expression.

• group2series - this table links data about experiment (series) and results of analysis (groups).
   

• function - sometimes genes can be grouped by function, for example overexpressed in G2. This function will be somehow related with GeneOntology.  

• group2function - this table will provide information about genes whose expression is changed and related with the specified function or state.
  Warning: this is some draft proposal that should be refined 
   

• TO DO - we need to design tables structure to store row microarray data. 

Publications

publications - stores publication references.
attachments - full text articles that can be associated with publication reference.

Publications table fields can be divides into 2 groups:

• MEDLINE fields - this fields corresponds to MEDLINE fields

• auxiliary fields - these fields allow user to customize this data - add custom key words, comments, importance. 

Automatic extraction data from MEDLINE

If publication has MEDLINE ID, then to add this publication to the publications table user should insert into the form only PUBMED ID (and possibly custom key words and comments) and all other information (title, authors, abstract, etc.) will be extracted from MEDLINE automatically.

Publications also can be categorized. We can develop the tree of categories and assign articles for one or several categories. This can simplify the publication references management.

Resources

The main purpose of this section is to build a list of useful resources. A resource can be link to the database, website, software, data format, etc.

• resources - list of resources

• resourceAttachments - arbitrary attachment. For example, for software resource it can be corresponding program installer.

Like publications resources also can be categorized.

Currently we have collected list of resources relevant to microarrays (microarray databases, supplementary information for the articles with microarray data, etc.)

Chemoinformatics

We can try to expand BioUML workbench so it will be able to cover main issues related with chemical substances and their pharmacological activities. Chemoinformatics section will contain additional tables to store information that can be stored as BioUML data.

Embedding chemoinformatics data into BioUML workbench

We can reuse concepts, substances, semantic relations and diagrams BioUML tables for formal description of pharmacological actions of different substances.

• concepts - it can be used for description of pharmacological activity.
  Currently BioUML defines 4 concept types:

  • activity - any biological or physiological function or activity (similar with GO definition)

  • process - any biological or physiological process (similar with GO definition)

  • state - state of organism or its part

  • concept - general type if we can classify the concept as activity, process or state.

  While sometimes is unclear whether the concept is activity or process, BioUML workbench allows customer to redefine concept types.
  Possibly we should define  some additional type for pharmacological activities.
   

• substances - it table can be used to describe pharmacological substances. For this purpose we will add some additional fields:

  • CAS - CAS Registry Number

  • Chemical Name - may be we can reuse completeName field for this purpose

  • Structure - structure in MOL format

  • formula 
     

• semantic relations - their can be used do describe relationships between different pharmacological activities
   

• substanceAction - more specialized type of semantic relation. It will be used for description of relationships between substances and concepts (pharmacological activities). Following fields will be added for this purpose:

  • IC50
  • ? something else ?
     

•  diagrams - we can try to use BioUML diagrams. Semantic network diagram type can be used for description of relationships between different pharmacological activities, physiological processes and influence of chemical substances on them. Biological pathway diagram type can be used for description of mechanism how pharmocological agent interacts with metabolic pathways or gene networks. If it will be necessary we can provide additional diagram types for presentation of chemoinformatics data.

Thus these chemoinformatics data will be available to user in different forms:

• manually created BioUML diagrams

• automatically generated graphs. BioUML workbech can automatically generate graphs that will show all relations between selected substances or pharmacological activity and other components.

• tables, available through Cyclonet web interface. We can develop a set of views for different purposes.

Additional tables

targets - contains list of possible targets. 
targetReferences - literature references for the targets.

Organizational issues

Blogs

The purpose of this section is to inform project participants and database users about current project progress. Current blogs implementation is something similar with forum or guest book, so it can be used for this purposes also. Any authorized project member can post new blog message or comment an existing blog.

• blogs - blog messages

• blog_comments - blog comments

Teams information

The purpose of these sections is to provide information about team members so any project member can easily get information about members of other teams, their contact information,  role in project, etc.

• person - information about person (project member)

• users, user_roles, roles - authorization information used by BeanExplorer. Any authorized project member should have private user name and user password. After authorization BeanExplorer will remember this information and can automatically fill fields: whoInserted___, whoModified___, creationDate___, modificationDate___.

• teams - information about our 4 teams. Currently information from grant proposal is used.

• team_persons - list of persons for each team.

• companies - information about institutes, companies and their departments. If company have mainCompanyID then it is considered as a department of main company.

• employments - information in which companies persons are working. A person can work in several companies simultaneously.

• Some auxiliary information (not completed yet):

  • bookmarks - bookmarks for companies and persons

  • addresses - company or person address

  • messages - e-mail messages associated with company or person.

  Issue Online Request for Proposal .