From: Egon W. <e.w...@sc...> - 2006-02-09 08:47:33
|
On Thursday 09 February 2006 09:24, nik...@no... wrote: > I just started playing around with CDK and during my first 'simple > learning examples' I also tested the calculation of surfaces on PDB files > ... > > Here I had two problems arising which might actually eb bugs (however, I > don't know exactly since I don't know if I am doing something wrong ;-) > > 1. When reading in PDB files one sometimes has files for which the RMEARK > line includes empty lines, i.e something like: > > REMARK 0 > REMARK 0 > REMARK 2 DATE OF DEPOSITION ... > REMARK 2 X-RAY DATA ... > REMARK 2 > REMARK 2 STRUCTURE SOLVED BY ... > > where often the 'spacefiller' REMARK 2 line has only 10 fields. when > reading these kind of PDB files PDBReader.java will throw an 'String index > out of range: -1' error since in line 263 of PDBReader field entry 11 is > tried to be accessed without any check. Can you send me the file, or give the PDB code if it is from the PDB db? > Now I was wondering if I am using > a wrong function or if this is just something which has not occured before > since most people don't use any PDB files?!? The PDB format is quite loose, and in some cases don't even stick to the specs, so problems due in either our code, or due to mismatches with specs are not uncommon. > 2. When calculating the numerical surface with NumericalSurface.java on a > read in PDB file I get a CDK exception : > org.openscience.cdk.exception.NoSuchAtomTypeException when the VDW radii > should be configured (line 318). I circumvented this by assigning the VDW > radii to the Protein as in the ProteinPocketFinder.java file (function > assignVdwRadiiToProtein) prior to calling NumericalSurface(). However, I > was wondering if there is an easy way to use the configVDW of the > NumericalSurface class? I have not used those code before, but I think it would be cleaner if NumericalSurface() would let the user do the VdW radius assignments. But I need to look at it... More later. Egon -- e.w...@sc... PhD student on Molecular Representation in Chemometrics Radboud University Nijmegen Blog: http://chem-bla-ics.blogspot.com/ http://www.cac.science.ru.nl/people/egonw/ GPG: 1024D/D6336BA6 |