Protein-DNA Interface Database: PDIdb
Wow. We’ve covered a good number of programs, databases and servers here on P212121, but I still get blown away by the great work researchers are putting out. We recently talked about the crystallization of protein-DNA complexes and wanted to follow up with a great database for those that are looking for structural information. The PDidb (Protein-DNA Interface Database) really owns the protein-DNA database landscape (if you are looking for others see the section in our crystallography programs list). So what makes this database special?
Inputs, the database contains 922 entries (they use a 2.5 A cut off from the structures in the PDB) that can be filtered extensively such as by whether the DNA strands have sticky ends, classifications based on interactions and of course simply using the PDB ID.
The database also has a statistics page with pie graphs that are click-able so that you can easily view the structures in that particular category.
The output is sweet. The upper left panel generates the PDB in JMol so that you can instantly examine the structure. The main panel contains various statistics of interest.
Finally and this is really nice if you need to get a presentation together with a simple click (on the pdf symbol below Jmol) – the structure can be outputted using nucplot to show the pertinent interactions.
My only compliant is that the single strand DNA results contain double stranded structures.
TomasN
June 4th, 2010 at 9:38 PM #
Hello Sean,
All DNA in the current version of PDIdb is double stranded. Double/single strand filter in the current version of the database refers to whether the DNA in the original crystal is single or double stranded. If your filter is single strand then you will select all complexes whose DNA is single stranded in the asymmetric unit of the crystal. In the next version of the database we will include single stranded DNA in the biological unit.
By the way, nice blog!!
All my best