PDB column overlap --PDB2PQR,APBS
Hi all, I am trying do Poisson-Boltzmann calculations through outside software, PDB2PQR and APBS. I am prepare my initial pdb file in chimera using the 'write pdb' method. The final structure is quite large, 737058 atoms, so I am getting overlaps in the data columns after I write the PDB file. This doesn't create a problem for rendering the PDB file but I get parsing errors and other strange occurrences when I try to create a complementary PQR file. Does anyone know of a work around? Here are links to my structures, if you look you will see that a significant portion of the structure is getting cut off in the PQR. I can't seem to generate more than about 99760 atoms, no matter what I try. http://dl.dropbox.com/u/42179401/2TMV.18H.pdb http://dl.dropbox.com/u/42179401/2TMV.18Ha.pqr Can anyone think of a potential fix? I've experienced this same problem with smaller helix pdb files. Thank you, Nikolay Rodionov
Hi Nikolay, I don't have an answer about how to make the steps work for huge numbers of atoms (especially given that PDB2PQR and APBS are not our own software -- did you also ask their developers?), but personally I would do Poisson-Boltzmann calculations on a single subunit only, without solvent, since the continuum is meant to model solvent. The ESP coloring for that subunit could be repeated on the other subunits with the "msc" command Tom Goddard mentioned in an earlier response. However, if your analysis demands having all the atoms for the whole system for some reason, my only other idea is to chop the data into pieces (write out PDBs for parts, convert to PQR files for those parts, then string together the PQR files into one giant PQR file). However, I have no idea whether you would have to manually renumber that PQR file, whether APBS would accept such a big PQR file, or if it does, whether it would take forever to run it. Best, Elaine ----- Elaine C. Meng, Ph.D. UCSF Computer Graphics Lab (Chimera team) and Babbitt Lab Department of Pharmaceutical Chemistry University of California, San Francisco On May 16, 2012, at 2:59 PM, Nikolay Igorovich Rodionov wrote:
Hi all,
I am trying do Poisson-Boltzmann calculations through outside software, PDB2PQR and APBS. I am prepare my initial pdb file in chimera using the ‘write pdb’ method. The final structure is quite large, 737058 atoms, so I am getting overlaps in the data columns after I write the PDB file. This doesn’t create a problem for rendering the PDB file but I get parsing errors and other strange occurrences when I try to create a complementary PQR file. Does anyone know of a work around?
Here are links to my structures, if you look you will see that a significant portion of the structure is getting cut off in the PQR. I can’t seem to generate more than about 99760 atoms, no matter what I try.
http://dl.dropbox.com/u/42179401/2TMV.18H.pdb http://dl.dropbox.com/u/42179401/2TMV.18Ha.pqr
Can anyone think of a potential fix? I’ve experienced this same problem with smaller helix pdb files.
Thank you, Nikolay Rodionov
Thank you Elaine. I want to visualize the force fields lines through the pore of the helix, I am going to try to isolate the amino acids (and a little extra to be safe) responsible for creating the pore's surface and just create helix based off of that data. I am hoping that the consequent surface charges inside the new pore will reflect the natural surface charge. Since you have mentioned sectioning off by subunits I want to tell you this in case some in the future needs help with a similar issue. Calculation results are significantly different based on how you treat a quandary protein structure, either as individual models or a single molecule. When I work use coulomb's law electrostatic coloring on a helix with individual surface calculations for each subunit I get a large negative potential (-10 eV) in localized regions surround Glu97 and Glu106 of each subunit. On the otherhand, when I combine all of the subunits into a single manageable helix molecule and do one large surface/ESP calculation for all of the atoms combined I get a large negative potential (-10eV) across the entire pore. I've attached images. I think that the single molecule approach give more accurate results. What do you think? Do you think that this calculation difference extend to other tools as well such as addh? Nikolay Rodionov -----Original Message----- From: Elaine Meng [mailto:meng@cgl.ucsf.edu] Sent: Wednesday, May 16, 2012 6:46 PM To: Nikolay Igorovich Rodionov Cc: chimera-users@cgl.ucsf.edu BB Subject: Re: [Chimera-users] PDB column overlap --PDB2PQR,APBS Hi Nikolay, I don't have an answer about how to make the steps work for huge numbers of atoms (especially given that PDB2PQR and APBS are not our own software -- did you also ask their developers?), but personally I would do Poisson-Boltzmann calculations on a single subunit only, without solvent, since the continuum is meant to model solvent. The ESP coloring for that subunit could be repeated on the other subunits with the "msc" command Tom Goddard mentioned in an earlier response. However, if your analysis demands having all the atoms for the whole system for some reason, my only other idea is to chop the data into pieces (write out PDBs for parts, convert to PQR files for those parts, then string together the PQR files into one giant PQR file). However, I have no idea whether you would have to manually renumber that PQR file, whether APBS would accept such a big PQR file, or if it does, whether it would take forever to run it. Best, Elaine ----- Elaine C. Meng, Ph.D. UCSF Computer Graphics Lab (Chimera team) and Babbitt Lab Department of Pharmaceutical Chemistry University of California, San Francisco On May 16, 2012, at 2:59 PM, Nikolay Igorovich Rodionov wrote:
Hi all,
I am trying do Poisson-Boltzmann calculations through outside software, PDB2PQR and APBS. I am prepare my initial pdb file in chimera using the 'write pdb' method. The final structure is quite large, 737058 atoms, so I am getting overlaps in the data columns after I write the PDB file. This doesn't create a problem for rendering the PDB file but I get parsing errors and other strange occurrences when I try to create a complementary PQR file. Does anyone know of a work around?
Here are links to my structures, if you look you will see that a significant portion of the structure is getting cut off in the PQR. I can't seem to generate more than about 99760 atoms, no matter what I try.
http://dl.dropbox.com/u/42179401/2TMV.18H.pdb http://dl.dropbox.com/u/42179401/2TMV.18Ha.pqr
Can anyone think of a potential fix? I've experienced this same problem with smaller helix pdb files.
Thank you, Nikolay Rodionov
Hi Nikolay, The figures look pretty nice -- I see you were able to display surfaces! It is not surprising the ESP results are different, comparing monomer to multimer. At least both images show some concentration of negative potential in the pore surface, and Coulombic ESP is meant for visualization (identification of more positive and negative regions in a given structure, or differences of surface potential between different but similar structures) rather than for absolute quantitative accuracy. I do agree that in your system, it is better to include the superstructure of multiple subunits in the ESP calculation, if possible, and that it would be really interesting to know if Poisson-Boltzmann ESP shows a similar enhancement of the potential in the superstructure! My second (chopping data) suggestion was putting the atoms back together in one file before running APBS, so everything would be in one ESP calculation, but it might not work. As to whether parts vs. whole affect calculations, I would expect yes for ESP calculations. For Add Charge on standard residues, no (since the charges for standard residues are just obtained from lookup tables). For AddH, the results either depend on all atoms open in Chimera at the time (if the "Consider each model in isolation" option is turned off) or all atoms within the same model but not other models (if that option is turned on). Best, Elaine ----- Elaine C. Meng, Ph.D. UCSF Computer Graphics Lab (Chimera team) and Babbitt Lab Department of Pharmaceutical Chemistry University of California, San Francisco On May 16, 2012, at 4:01 PM, Nikolay Igorovich Rodionov wrote:
Thank you Elaine. I want to visualize the force fields lines through the pore of the helix, I am going to try to isolate the amino acids (and a little extra to be safe) responsible for creating the pore's surface and just create helix based off of that data. I am hoping that the consequent surface charges inside the new pore will reflect the natural surface charge.
Since you have mentioned sectioning off by subunits I want to tell you this in case some in the future needs help with a similar issue. Calculation results are significantly different based on how you treat a quandary protein structure, either as individual models or a single molecule. When I work use coulomb's law electrostatic coloring on a helix with individual surface calculations for each subunit I get a large negative potential (-10 eV) in localized regions surround Glu97 and Glu106 of each subunit. On the otherhand, when I combine all of the subunits into a single manageable helix molecule and do one large surface/ESP calculation for all of the atoms combined I get a large negative potential (-10eV) across the entire pore.
I've attached images. I think that the single molecule approach give more accurate results. What do you think? Do you think that this calculation difference extend to other tools as well such as addh?
Nikolay Rodionov
Hi I am looking for a command line version of the CENTER command in VOLUME VIEWER. Can anybody please give me a hint! I'd like to use this command in a script to re-center a volume. I have also tried to the use the script radius_of_gyration.py but there is something I do not understand: I load the script with 'open file' but then nothing happens. Where is the output directed to? How could I use the output to re-center my volume? Thanks, Dieter ------------------------------------------------------------------------ Dieter Blaas, Max F. Perutz Laboratories Medical University of Vienna, Inst. Med. Biochem., Vienna Biocenter (VBC), Dr. Bohr Gasse 9/3, A-1030 Vienna, Austria, Tel: 0043 1 4277 61630, Fax: 0043 1 4277 9616, e-mail: dieter.blaas@meduniwien.ac.at ------------------------------------------------------------------------
Hi Dieter, I'm not sure if there is a command that does exactly the same thing as the Center button in Volume Viewer, but for a similar result you might try the "window" command: <http://www.cgl.ucsf.edu/chimera/docs/UsersGuide/midas/window.html> The radius_of_gyration.py script is for atomic coordinates, not volume data, at least if you are talking about the script that was posted on the list a couple of months ago: <http://plato.cgl.ucsf.edu/pipermail/chimera-users/2012-February/007257.html> It reports center of mass and radius of gyration values in the Reply Log (under Favorites in the Chimera menu). It doesn't move anything. To calculate the center of mass of a density map (also not moving anything), you could use the command "measure center": <http://www.cgl.ucsf.edu/chimera/docs/UsersGuide/midas/measure.html#center> That gives the center in grid indices of your map rather than in x,y,z coordinates. I hope this helps, Elaine ---------- Elaine C. Meng, Ph.D. UCSF Computer Graphics Lab (Chimera team) and Babbitt Lab Department of Pharmaceutical Chemistry University of California, San Francisco On May 17, 2012, at 1:34 AM, Dieter Blaas wrote:
Hi I am looking for a command line version of the CENTER command in VOLUME VIEWER. Can anybody please give me a hint! I'd like to use this command in a script to re-center a volume. I have also tried to the use the script radius_of_gyration.py but there is something I do not understand: I load the script with 'open file' but then nothing happens. Where is the output directed to? How could I use the output to re-center my volume? Thanks, Dieter
Hi Elaine, I was wondering if it is possible to chop up a large PDB file into numerous chains in chimera. That way the atom numbering should restart and I won't have parsing errors. Thank you, Nikolay Rodionov -----Original Message----- From: Elaine Meng [mailto:meng@cgl.ucsf.edu] Sent: Wednesday, May 16, 2012 6:46 PM To: Nikolay Igorovich Rodionov Cc: chimera-users@cgl.ucsf.edu BB Subject: Re: [Chimera-users] PDB column overlap --PDB2PQR,APBS Hi Nikolay, I don't have an answer about how to make the steps work for huge numbers of atoms (especially given that PDB2PQR and APBS are not our own software -- did you also ask their developers?), but personally I would do Poisson-Boltzmann calculations on a single subunit only, without solvent, since the continuum is meant to model solvent. The ESP coloring for that subunit could be repeated on the other subunits with the "msc" command Tom Goddard mentioned in an earlier response. However, if your analysis demands having all the atoms for the whole system for some reason, my only other idea is to chop the data into pieces (write out PDBs for parts, convert to PQR files for those parts, then string together the PQR files into one giant PQR file). However, I have no idea whether you would have to manually renumber that PQR file, whether APBS would accept such a big PQR file, or if it does, whether it would take forever to run it. Best, Elaine ----- Elaine C. Meng, Ph.D. UCSF Computer Graphics Lab (Chimera team) and Babbitt Lab Department of Pharmaceutical Chemistry University of California, San Francisco On May 16, 2012, at 2:59 PM, Nikolay Igorovich Rodionov wrote:
Hi all,
I am trying do Poisson-Boltzmann calculations through outside software, PDB2PQR and APBS. I am prepare my initial pdb file in chimera using the 'write pdb' method. The final structure is quite large, 737058 atoms, so I am getting overlaps in the data columns after I write the PDB file. This doesn't create a problem for rendering the PDB file but I get parsing errors and other strange occurrences when I try to create a complementary PQR file. Does anyone know of a work around?
Here are links to my structures, if you look you will see that a significant portion of the structure is getting cut off in the PQR. I can't seem to generate more than about 99760 atoms, no matter what I try.
http://dl.dropbox.com/u/42179401/2TMV.18H.pdb http://dl.dropbox.com/u/42179401/2TMV.18Ha.pqr
Can anyone think of a potential fix? I've experienced this same problem with smaller helix pdb files.
Thank you, Nikolay Rodionov
Hi Nikolay, See "split" command: <http://www.cgl.ucsf.edu/chimera/docs/UsersGuide/midas/split.html> Elaine On May 17, 2012, at 9:58 AM, Nikolay Igorovich Rodionov wrote:
Hi Elaine,
I was wondering if it is possible to chop up a large PDB file into numerous chains in chimera. That way the atom numbering should restart and I won't have parsing errors.
Thank you, Nikolay Rodionov
Or alternatively, can I break one large molecule into several models? -----Original Message----- From: Elaine Meng [mailto:meng@cgl.ucsf.edu] Sent: Wednesday, May 16, 2012 6:46 PM To: Nikolay Igorovich Rodionov Cc: chimera-users@cgl.ucsf.edu BB Subject: Re: [Chimera-users] PDB column overlap --PDB2PQR,APBS Hi Nikolay, I don't have an answer about how to make the steps work for huge numbers of atoms (especially given that PDB2PQR and APBS are not our own software -- did you also ask their developers?), but personally I would do Poisson-Boltzmann calculations on a single subunit only, without solvent, since the continuum is meant to model solvent. The ESP coloring for that subunit could be repeated on the other subunits with the "msc" command Tom Goddard mentioned in an earlier response. However, if your analysis demands having all the atoms for the whole system for some reason, my only other idea is to chop the data into pieces (write out PDBs for parts, convert to PQR files for those parts, then string together the PQR files into one giant PQR file). However, I have no idea whether you would have to manually renumber that PQR file, whether APBS would accept such a big PQR file, or if it does, whether it would take forever to run it. Best, Elaine ----- Elaine C. Meng, Ph.D. UCSF Computer Graphics Lab (Chimera team) and Babbitt Lab Department of Pharmaceutical Chemistry University of California, San Francisco On May 16, 2012, at 2:59 PM, Nikolay Igorovich Rodionov wrote:
Hi all,
I am trying do Poisson-Boltzmann calculations through outside software, PDB2PQR and APBS. I am prepare my initial pdb file in chimera using the 'write pdb' method. The final structure is quite large, 737058 atoms, so I am getting overlaps in the data columns after I write the PDB file. This doesn't create a problem for rendering the PDB file but I get parsing errors and other strange occurrences when I try to create a complementary PQR file. Does anyone know of a work around?
Here are links to my structures, if you look you will see that a significant portion of the structure is getting cut off in the PQR. I can't seem to generate more than about 99760 atoms, no matter what I try.
http://dl.dropbox.com/u/42179401/2TMV.18H.pdb http://dl.dropbox.com/u/42179401/2TMV.18Ha.pqr
Can anyone think of a potential fix? I've experienced this same problem with smaller helix pdb files.
Thank you, Nikolay Rodionov
participants (3)
-
Dieter Blaas -
Elaine Meng -
Nikolay Igorovich Rodionov