Dear all,
We are happy to announce the PDB_REDO web server at 
        http://xtal.nki.nl/pdb_redo
The
 PDB_REDO server refines any structural model you are working on, by 
optimising REFMAC refinement and by partial model rebuilding. The 
process is fully automated [1] so you only need to provide an initial 
structure model and your reflection data. The server will give you a new
 structure model, MTZ files for new maps, some validation data, 
optimised parameters for REFMAC, and Coot scripts to walk you through 
the structural changes.
Robbie and Tassos

Many thanks to:Fei
 Long and Garib Mushudov for donating web server code and helping with 
customization, and rather obviously for REFMAC development.Paul Emsley for the 
Coot script help and guidelines.Krista Joosten, Serge Cohen and Gert Vriend and 
other development contributors.NWO/CW for paying Robbie's salary and the CPUs
Summary of key features:Extensive
 refinement parametrisation for Refmac. This includes geometrical and B 
factor restraint weighting, dealing with NCS, twining and TLS, B-factor 
model  selection through the Hamilton R ratio test [2], and resolution 
cut-off determination through paired refinement [3]. All refinement settings 
are made available for reading them back in the REFMAC CCP4i gui.Optimisation 
of side chain conformations and peptide bond orientations  [4].Thorough
 validation, both in terms of fit to the data and geometric quality, 
with special attention to ligands and their interactions [5]. New "boxplot" 
statistics showing how good your model is compared to the PDB and the PDB_REDO 
"nearest resolution range" models.Easy-to-interpret description of model 
changes such as colour-coded validation scores and real-space correlation 
plots.A script that  guides you through changed model features (rotamers, 
flips, completed residues, deleted waters) in Coot.
Data safety:
A
 login ID is created on the fly and allows easy job management. Your 
data are thus password protected, and can be viewed and deleted only by 
you (well, also by Robbie and Tassos, to be honest) directly after 
your job is finished. If you do not delete jobs, we will, after 
approximately 30 days (depending on the success of the service). After 
this, only statistical data (no sequences or coordinates) are retained. 
These said, we expect that Mr. Snowden and associates and/or the NSA can
 make their way and access your data. We make no implicit  or explicit 
guarantees for data safety.
Licensing:
The
 PDB_REDO server is free to all academic and commercial users who have a
 CCP4 license. We do not make checks, we do not ask you to click buttons
 so we can sue you later on, but we (us and the CCP4 team) simply expect
 you to behave properly.  The developers of YASARA and FoldX have  waved
 any licence requirements for their software on the server.
Hardware:
We
 have made available a 4-CPU 6-core machine, giving access to 48 
parallel processes using hyper-threading. Jobs are queued on a 
first-come first-served basis, when necessary. Jobs can take from a few 
minutes up to several hours depending on structure size and resolution. 

References:
[1]
 Joosten RP, Joosten K, Murshudov GN, Perrakis A. PDB_REDO: constructive
 validation, more than just looking for errors. Acta Cryst. 2012; 
D68:484-496 http://dx.doi.org/10.1107/S0907444911054515

[2] Merritt EA. To B or not to B: a question of resolution? Acta Cryst. 2012; 
D68:468-477 http://dx.doi.org/10.1107/S0907444911028320

[3] Karplus PA, Diederichs K. Linking crystallographic model and data quality. 
Science 2012; 336:1030-1033 http://www.sciencemag.org/content/336/6084/1030.full

[4]
 Joosten RP, Joosten K, Cohen SX, Vriend G, Perrakis A. 
Automatic rebuilding and optimization of crystallographic structures in 
the 
Protein Data Bank. Bioinformatics 2011; 27:3392-3398.  
http://dx.doi.org/10.1093/bioinformatics/btr590

[5]
 Cereto-Massagué A, Ojeda MJ, Joosten RP, Valls C, Mulero M, Salvado MJ,
 Arola-Arnal A, Arola L, Garcia-Vallvé S, Pujadas G. The good, the 
bad and the dubious: VHELIBS, a validation helper for ligands and binding 
sites. J Cheminform. 2013; 5(1):36. http://dx.doi.org/10.1186/1758-2946-5-36

                                          

Reply via email to