Thank you for the quick response. Randy Read wrote:
On the point of "tuning" parameters, I'm not sure what you mean. In a particular case, you should know what you put into your
This is probably a good point at which to mention that I am not a crystallographer or structural biologist, but a computational scientist working on parameter sweep mechanisms to tune the molecular replacement phase of structure discovery. I have spoken with structural biologists working in industry who say they do exactly kind of parameter tuning and parameter sweep -- rather than assume they know more about their crystal/structure properties than they think. I also find in literature "hints" such as Figure 3 in this December 2007 Acta D paper on MR (using CCP4 Molrep, not Phaser): http://www.ncbi.nlm.nih.gov/pmc/articles/PMC2394805/?tool=pubmed that suggest the underlying algorithms do not rely on parameter settings matching "reality" to produce the most promising results -- note in Figure 3 a 0.2 and 0.9 completeness give a similar Rfree result. Of course, a priori structure/crystal knowledge would probably allow correct selection from the various peaks, but it is still relevant that the underlying algorithms in molrep produce similar strong results with dramatically different input parameters.
2. Given that the potential solutions in the .sol file are sorted by LLG, I'm not sure where the idea would come from that they could be given in the order they were found. You can follow the solutions in
What confused me here was that the same file is being repeatedly over-written as Phaser progresses.
4. If Phaser reports that there is no scattering in a model, it means that you have supplied an empty PDB file, or one where all the
It turns out this was a bug on our part -- In a small number of cases (about 2%) our PDB trimming algorithm was inserting bogus END entries into the intermediate PDB file. Phaser (rightly) stopped processing when it encountered these, and produced the given error. Other tools in our processing chain were more forgiving, so it appeared to be a Phaser problem. Ian -- Ian Stokes-Rees W: http://sbgrid.org [email protected] T: +1 617 432-5608 x75 SBGrid, Harvard Medical School F: +1 617 432-5600