Hi, What I meant is that the solution that showed the TFZ-equivalent score for the second component should have been the one with the highest LLG after adding the second component, and it wasn't in your example. That's why I thought you must have made it up.
Randy On 31 Jul 2013, at 12:30, Chen Zhao <chenzhaoh...@gmail.com> wrote: > Dear Prof. Read, > > Thank you for your detailed explaination. I think now I am clear about what > each set of scores represent. But what do you mean by what the .sol should > show? I did "make it up" by selecting two solutions sets from two different > runs, and I am sorry for not making it clear at the begining. > > Thank you so much! > > Sincerely, > Chen > > > On Wed, Jul 31, 2013 at 6:35 AM, Randy Read <rj...@cam.ac.uk> wrote: > Dear Chen, > > For each component that is placed, Phaser reports the Z-scores for the > rotation function (RFZ) and translation function (TFZ), along with the number > of packing clashes (PAK) at this point and the LLG. For the top solution at > that point in the search, Phaser also reports the TFZ-equivalent (TFZ=) > score, which is what the Z-score would have been for the translation search > if the refined orientation from the rigid-body refinement had been used. We > introduced the TFZ-equivalent to eliminate the dependence of the TFZ on the > quality of the orientation used for the particular translation search that > (after refinement) ended up giving the best solution. However, it takes some > CPU time to compute a random sampling of translations for that orientation, > so we only do it for one solution after adding each component. > > Your .sol file doesn't quite make sense in terms of what it should show. Did > you make it up to illustrate your point? What you might see is that, for the > top solution, the first LLG value (from placing the first component) is lower > the the corresponding LLG value for a solution lower in the final list. > However, you might see a TFZ== entry immediately after the first LLG for > whichever solution had the highest LLG at that point in the search. If a > solution that was worse after placing the first component ends up with the > top LLG after adding the second component to that solution, then there will > be a TFZ== entry after the second LLG value (corresponding to the LLG after > placing the second component. > > I hope that helps! There's also an explanation on our Phaser Wiki: > http://www.phaser.cimr.cam.ac.uk/index.php/Molecular_Replacement#Annotation, > which gets into some even more obscure aspects of this annotation line. > > Randy Read > > On 31 Jul 2013, at 01:24, Chen Zhao <chenzhaoh...@gmail.com> wrote: > >> Dear all, >> >> I am sorry for my new-comer question, but I am not clear what the multiple >> RFZ, TFZ, PAK and LLG scores belong to after each solution set in the phaser >> output (eg. .sol) file. My bet is that each of them corresponds to each of >> the solutions listed below, but I am not sure. Also I know that TFZ== >> represents the TFZ score based on the refined orientation, but why does it >> sometimes appear in the middle instead of at the end? >> >> One example could look like this: >> >> SOLU SET RFZ=5.4 TFZ=3.6 PAK=0 LLG=67 TFZ==1.8 RFZ=2.2 TFZ=5.1 PAK=3 LLG=91 >> LLG=102 >> SOLU SPAC XXX >> SOLU 6DIM ENSE ensemble1 EULER X X X FRAC X X X BFAC X >> SOLU 6DIM ENSE ensemble1 EULER X X X FRAC X X X BFAC X >> SOLU ENSE ensemble1 VRMS X >> SOLU ENSE ensemble2 VRMS X >> >> SOLU SET RFZ=4.2 TFZ=2.8 PAK=0 LLG=51 RFZ=3.1 TFZ=5.1 PAK=1 LLG=71 LLG=86 >> TFZ==5.3 >> SOLU SPAC XXX >> SOLU 6DIM ENSE ensemble1 EULER X X X FRAC X X X BFAC X >> SOLU 6DIM ENSE ensemble1 EULER X X X FRAC X X X BFAC X >> SOLU ENSE ensemble1 VRMS X >> SOLU ENSE ensemble2 VRMS X >> >> Thank you so much! >> >> Sincerely, >> Chen > > ------ > Randy J. Read > Department of Haematology, University of Cambridge > Cambridge Institute for Medical Research Tel: + 44 1223 336500 > Wellcome Trust/MRC Building Fax: + 44 1223 336827 > Hills Road E-mail: rj...@cam.ac.uk > Cambridge CB2 0XY, U.K. www-structmed.cimr.cam.ac.uk > > ------ Randy J. Read Department of Haematology, University of Cambridge Cambridge Institute for Medical Research Tel: + 44 1223 336500 Wellcome Trust/MRC Building Fax: + 44 1223 336827 Hills Road E-mail: rj...@cam.ac.uk Cambridge CB2 0XY, U.K. www-structmed.cimr.cam.ac.uk
