Menu
▾
▴
Re: [PyMOL] Fw: symmetry and pdb remark
|
From: Spencer B. <spe...@gm...> - 2017-02-03 20:21:52
|
Smith, Please reply on the list so that others can benefit from the responses. First, remember that the lattice is the "correct" crystal structure, and so all the choices of unit cell are equally valid mathematically. So if you're hoping for one particular arrangement you have to think about what your goal is. Personally, I prefer the '2zao_6mates' arrangement (a compact left-handed helix around the C axis). The paper seems to indicate that the full-length protein would form hexamers. While there must be some conformational rearrangement relative to the current structures (otherwise they would aggregate), it seems plausible that the hexameric interface might be similar to the large interface (>1000Å^2 <http://www.eppic-web.org/ewui/#id/2zan>) shown here. I've found that it generally requires some manual intervention to choose the most aesthetically pleasing and scientifically clear arrangement for the unit cell. Typically my pymol workflow goes something like this: - Generate a 2x2 or 3x3 layer of unit cells using the supercell command: - supercell 2,2,1, 2zan - Visually inspect the lattice and determine what the most interesting decomposition is - disable objects until you have just what you want to keep. I usually do this in the GUI, but in this case the result is - disable *; enable m000_2 m010_1 m100_3 m110_4 m110_5 m110_6 supercell - For viewing unit cells, it is often useful to use orthoscopic perspective: - set orthoscopic, on You also mention obliquely pymol's symexp command. This is intended for generating something like a sphere of symmetry partners around the asymmetric unit. The distance parameter refers to the distance from some atom of the query selection to an atom of the symmetry mate. Usually 5-10Å is used. The command is not really suitable for generating unit cells. -Spencer On Fri, Feb 3, 2017 at 5:41 PM, Smith Liu <smi...@16...> wrote: > Dear Spencer, > > > > Thanks for your reply on my question "symmetry and pdn remarks". Here I > have more questions related to the issue. > > > > Attached 1 was the 6 mates coordinates for 2zao, and attached 2 was the 6 > mates coordinates for 2zn. I prepared the mate coordinates based on the > following: > > > > Use Coot. > > Draw->Cell&Symmetry->Master switch No->show unit cell yes > > > > Then > > > > Extensions->Modelling->New Molecule from operator->cut and paste the 5 > non-identity operators from Rupp's space group applet into the X,Y,Z > window. repeat for all 5. > > > > The arrangement for the 6 mates for both 2zao and 2zan was exactly same as > the what Chimera produced with its "pack in Unit Cell" function unclicked (you > may be not familiar with Chimera, but this will not impede you understand > this e-mail) > > > > If you view my first 2 attached files, you will find the arrangements for > the 6 mates of 2zao and 2zan were absolutely different. Thus 1 attached > file (1st one or 2rd one), as mainly prepared by Coot, was wrongly > processed by Coot, or the original file , either the pdb for 2zan or for > 2zao, has misled coot. Will you please explain how the pdb file has mislead > coot? > > > > As for the first 2 attached files indicated significant difference of 6 > mates arrangement for 2zao and 2zan, then will you please advise based on > what we decide which attachment was the correct mate files? > > > > For the attachment 3, it was the 6 mates coordinate of 2zao prepared by > Chimera with its "pack in Unit Cell" function clicked, and as introduced by > Chimera, the function to click "pack in Unit Cell" was to " pack the chains > so that their centers fall within one unit cell box".My comment: The mates > produced by Chimera with its "pack in Unit Cell" function clicked, was > something like the mates produced by Pymol, but seems arrangement not > identical as viewed by naked eye). > > > > Then for attachment 1 and attachment 3, which can represent the molecule > arrangement in the real crystal? If both attachment 1 and attachment 3 > represent the molecule arrangement in the real crystal, then will you > please expand attachment 3 as littile as possible, so that a new pdb file > with the least number of molecules will be produced, which contained both > the information in attachment 1 and attachment 3? > > > Next, Pymol can generate symmetry mates within a specific distance (4A, > 5A, 6A,..). Will you please explain to me the definition of that distance? > > > Fourth, based on the symetry mates generated by Pymol, do you think > whether we can view the mate arrangement as helix 1? Or do the symmetry > mates generated by Pymol was in "pack" arrangement as the "pack" > arrangement produced by Chimera? > > > > > I am looking forward to getting a reply from you. > > > > Smith > > > > > > > At 2017-02-03 17:28:51, "Spencer Bliven" <spe...@gm...> wrote: > > This isn't even a difference in choice of origin, but rather the choice of > asymmetric unit. 2zam chose an asymmetric unit near the y axis, while 2zam > chose an asymmetric unit one cell to the right. You can see this by > inspecting the alignment matrix to put 2zam onto 2zan. > > >align 2zam, 2zan > >print cmd.get_object_matrix("2zam") > (0.9898772835731506, -0.007360012270510197, -0.1417350471019745, > 76.22661536234315, > 0.007859288714826107, 0.999964714050293, 0.002963125705718994, > -7.941656963507889, > 0.14170823991298676, -0.004047067370265722, 0.9899001717567444, > 7.548435961979237, > 0.0, 0.0, 0.0, 1.0) > > This is quite close to the identity matrix plus a 76Å translation along x > (a = 75.61 for 2zan). > > Chimera seems to display a unit cell such that the centroid of the > asymmetric unit is located within that unit cell. The equivalent pymol > command (the supercell <https://pymolwiki.org/index.php/Supercell> > plugin) generates the unit cell relative to the origin. Thus in pymol the > two unit cells overlap perfectly: > > > >supercell 1,1,1,2zan,red,name=cell_2zan,prefix=n > >supercell 1,1,1,2zam,blue,name=cell_2zam,prefix=m > >color red, n*; color blue, m* > > [image: Inline image 1] > > -Spencer > > On Thu, Feb 2, 2017 at 5:12 PM, Robert Campbell < > rob...@qu...> wrote: > >> Hello Smith, >> >> I had a look at the two structures. In some crystallographic space >> groups there are alternate choices for the origin that are all equivalent. >> It appears that this is what you are seeing here. >> >> If you look carefully at the mates generated separately for 2zam and >> 2zan, they are equivalent in their position and orientation relative to the >> original coordinates for each structure. There is simply a translational >> shift along two axes that relates the coordinates of 2zam and its mates to >> 2zan and its mates. >> >> There is a slight difference in the unit cell dimensions as well, which >> could make very small differences in the packing contacts. >> >> Cheers, >> Rob >> >> On Thu, 2017-02-02 12:37 +0800, Smith Liu <smi...@16...> wrote: >> >> > Dear All, >> > >> > >> > Here I make my question much clear. >> > >> > >> > For both PDB 2zan and 2zam, they are for the same protein, they >> > conformation were similar except that 2zam was apo and 2zam was ATP >> > binding. 2zaz was got by soaking the 2zam crystal with ATP. Both were >> > P65 space group >> > >> > >> > However by getting the mate coordinates of 2zam and 2zan by Coot or >> > Chimera, you will find their 6 mates arrange differently ( viewed by >> > Chimera or by Pymol). For 2zan, the 6 mates were separated with each >> > other, but for 2zam the 6 mates were connected. Can you explain to me >> > why their 6 mates arrange differently? >> > >> > >> > What is more, by pymol I align 2zan onto 2zam, and I get the PDB for >> > 2zan fitted to 2zam (no any text information from the pdb file left >> > after pymol saving, except the coordinates). Then I add all the text >> > information from 2zan (except the coordinates) to the PDB for 2zan >> > fitted to 2zam, and then I get the mate coordinates by Chimera or >> > Coort and then view the mates got by Chimera or Coot for the text >> > information added PDB for 2zan fitted to 2zam, I find the 6 mates >> > arrange like 2zam, rather like 2zan (viewed by Chimera). Thus, will >> > you please explain which remark information decide the mate >> > arrangement, as in the remark added PDB for 2zan fitted to 2zam? Why >> > after pymol alignment, the same remark information leads to different >> > mates arrangement? >> > >> > >> > I am looking forward to getting a reply from you. >> > >> > >> > Smith >> > >> > >> > >> > >> > >> > >> > >> > -------- Forwarding messages -------- >> > From: "Smith Liu" <smi...@16...> >> > Date: 2017-02-02 12:06:16 >> > To: pym...@li... >> > Subject: symmetry and pdb remark >> > >> > Dear All, >> > >> > >> > >> > I have a symmetry problem, which I hope I can get your help. >> > >> > >> > >> > For both PDB 2zan and 2zam, they are for the same protein, they >> > conformation were similar except that 2zam was apo and 2zam was ATP >> > binding. 2zaz was got by soaking the 2zam crystal with ATP. Both were >> > P65 space group >> > >> > >> > However by getting the mates of 2zam and 2zan, you will find their 6 >> > mates arrange differently ( viewed by Chimera). Can you explain to me >> > why their 6 mates arrange differently? >> > >> > >> > What is more, by pymol I align 2zan onto 2zam, and I get the PDB for >> > 2zan fitted to 2zam (no any remark information left after pymol >> > saving). Then I add all the remark information from 2zan to the PDB >> > for 2zan fitted to 2zam, and then I view the mates for the remark >> > added PDB for 2zan fitted to 2zam, I find the 6 mates arrange like >> > 2zam, rather like 2zan (viewed by Chimera). >> > >> > >> > Thus, will you please explain which remark information decide the >> > mate arrangement, as in the remark added PDB for 2zan fitted to >> > 2zam? Why after pymol alignment, the same remark information leads to >> > different mates arrangement? >> > >> > >> > I am looking forward to getting a reply from you. >> > >> > >> > Smith >> >> >> >> -- >> Robert L. Campbell, Ph.D. >> Adjunct Assistant Professor >> Dept. of Biomedical & Molecular Sciences, Botterell Hall Rm 644 >> Queen's University, Kingston, ON K7L 3N6 Canada >> Tel: 613-533-6821 >> <rob...@qu...> http://pldserver1.biochem.queensu.ca/~rlc >> >> ------------------------------------------------------------ >> ------------------ >> Check out the vibrant tech community on one of the world's most >> engaging tech sites, SlashDot.org! http://sdm.link/slashdot >> _______________________________________________ >> PyMOL-users mailing list (PyM...@li...) >> Info Page: https://lists.sourceforge.net/lists/listinfo/pymol-users >> Archives: http://www.mail-archive.com/pym...@li... >> > > > > > |
