Searching the MDB

Structure of the MDB

Available search interfaces

At the present we have 3 web-based interfaces to the MDB: interactive, forms-based, and SQL.

Interactive interface

A frames based interface that allows interactive searching and display of the results, via the CCMS/TSRI java-based macromolecular viewer.

This interface uses a frame-based document making use of Java and JavaScript, along with regular CGI forms support. This means that to use this interface to its full, you will need a browser that supports this features. Most of the modern and recent web browsers should not have problems in this respect.

This interactive interface allows the manipulation of results using our java Macromolecular Viewer. This viewer is stable and fast in the rendering of molecular structures. (Note: I've been able to run this over a PPP connection, using a 28.8kbps modem, under Windows 3.1 with Internet Explorer 3.0, in a laptop with a 486SX/333MHz and 8MB RAM (260 MB HD, 12MB free). It took about 3-4 minutes to load, but it run well afterwards. So I think it is stable --- Jesus M. Castagnetto).

• Using the query form in the left part of the search page, you will be able to select the criteria you want to use. Every time you make a change to this form, this will be reflected in the message frame (middle part of the right part of the page). In the message frame, the information about the current options, query status and results will be displayed.
• In case your query returns too many results you will always have the option to discard it and start over. Once the results have been processed, they will be displayed in the Macromolecule Model Manager, using which you can selectively view/hide/delete one or more of the models. A handy "Cummulative query" tab is available there, so you can alway access previous query results (provided they were not deleted, which discards the model).
• All the viewing and manipulations (distance/angle/dihedral measurements, rotation/translation/zoom, etc.) will be performed by using the Java applet viewer (which is why the search pages take some time to load). All this operations can be done in real time as the server connections are kept to a minimum by using a good dosis of Javascript and Java.
• If some quick help is needed, use the appropriate links in the graphic (or the text) navigation bar at the top (bottom) of the document.

HTML forms interface

You can also use our straight forms interface to search the MDB. This advanced queries can be made by using one of three forms, which increase in complexity (and thus, in flexibility).

The results of the queries are returned as a numbered list, with links allow downloading of the VRML 2.0 or PDB formatted structure files corresponding to the metal site, an option to see the metal site using our MacroMolecular viewer on a separate window (java support required), and links to the web page of the corresponding protein at the PDB and the "Molecules R Us" (NIH) sites. A short description is also presented.

First form:

Allows you to search for metal sites, by entering one or more PDB ids. Multiple identifiers are entered as a comma separated list, e.g.: 1hca,2sod

Second ("Simplified") form:

Allows searching for metal sites fitting a series of constrains:

• metal(s) - comma separated list (e.g.: zn,cu,ca)
• number of ligands - e.g. 5,6,8 (means "5 OR 6 OR 8 ligands")
• upper limit for resolution - e.g. 2.5 means "resolution <= 2.5 Å"
• upper limit for the r-value - e.g. 0.3 means "r-value <= 0.3"
• source of the experimental data - x-ray, nmr, etc.
• (partial) author name match - e.g. smit would match authors named: Smith, Smitters, Alsmither, etc.

And you will also be able to restrict how many results are displayed on one page, e.g. if you select to display 10 results per page (the default), and your query finds 26 metal sites, you will be navigate through 3 pages of results.

Third ("Advanced") form:

This forms allows you to search for metal sites using a more complete sets of constrains:

• metal(s) - comma separated list (e.g.: zn,cu,ca)
• number of ligands - e.g. 6,7 (means "6 OR 7 ligands"). And the ligands can be select further by:
  • indicating the number of ligands that need to be of a particular type (amino acids, nucleic acids, waters)
  • indicate that at least one of the protein ligands has to match one of a list of aminoacids
  • restricting the metal-ligand distance, by using a range notation, e.g. -1.6 means "<= 1.6 Å"; 1.8- means ">= 1.8 Å" and 1.7-2.4 will mean "between (and including) 1.7 and 2.4 Å"
• revision/deposition date of the structure - in a range notation, e.g. -1996, 1992-1998, etc.
• resolution limits - in a range notation, e.g. 2.5-, 1.8-2.2, etc.
• r-value limits - in a range notation, e.g. -0.3, 0.4-
• source of the experimental data - x-ray, nmr, etc.
• (partial) author name match - e.g. mor would match authors named: Morris, Elmore, etc.

As in the table before you can select how many results to show per page.

SQL interface

For those who like more flexibility, not only in the query terms but also on the display of results, we are making available a "raw" SQL interface to the MDB. No SQL syntaxis checking is made on the query, and the only restriction is that only simple SELECT statements are allowed. No sub-queries or aggregate statements can be used because the SQL engine we are using does not support it.(*)

Information on the structure of the searchable tables and columns in the MDB can be found in the search page, as well as a series of (explained) examples of SQL queries.

The results of the queries are presented in simple table form. No checking for the number of rows in the table is made, so if your query generates too many hits, you will end up with a very long table.

(*) We are not using a more powerful SQL engine because the database server and the web server are running on the same, shared machine.

What else can I access here?

If you have a question that is not answered here or in the help available in the search page, check our FAQ page. If you still do not find an answer to your inquiry, send e-mail to metallodb@scripps.edu

In case you are interested in the data we have generated during the automation process, check our (unwieldly named) Packages page. This page contains the files we have used in the production of our "raw" database. See the descriptions there for more information.