Re: [gmx-users] Reading XTC files from fortran90
Hi Jones, I can already do that for .xtc files. Seems that, unfortunatelly, the least odd way is to find a way to get the last pieces of information needed (atomic charges and masses) from the .tpr files. Other ways would include lots of file translations, or adaptions in the codes themselves which would lead to non-generality of it. I'd say the easiest is to get into gmx_editconf and look at the option -grasp or -mead which put the atomic charge and vdw radius in the b-factor and occupancy field of a .pdb file when given a .tpr file. It's trivial to make it write charges and masses instead. Would also be a good first exercise for C/Gromacs :p Sorry for not noticing the box shape in the .xtc file. I passed over that integer in the reading, dismissing it as only some kind of old feature useless now. Really sorry for that point. I'm not sure what you mean with shape and that integer. There's nothing much redundant in the .xtc file and there are no integers having to do with box shape. In fact, there is no shape... Cheers, Tsjerk -- Tsjerk A. Wassenaar, Ph.D. Junior UD (post-doc) Biomolecular NMR, Bijvoet Center Utrecht University Padualaan 8 3584 CH Utrecht The Netherlands P: +31-30-2539931 F: +31-30-2537623 ___ gmx-users mailing listgmx-users@gromacs.org http://www.gromacs.org/mailman/listinfo/gmx-users Please search the archive at http://www.gromacs.org/search before posting! Please don't post (un)subscribe requests to the list. Use the www interface or send it to [EMAIL PROTECTED] Can't post? Read http://www.gromacs.org/mailing_lists/users.php
Re: [gmx-users] Reading XTC files from fortran90
Jones de Andrade wrote: Hi Mark. Wait a second, you mean that editconf would allow me to have both charges and masses on a .pdb file directly from command line (having other matters asking for my presence absolutelly now, so please forgive me if this message becomes somehow strange) No, nobody said that. Please read free advice carefully, since there's no better way to annoy the giver than to not be understood when they've expressed an idea clearly. So, I would have atom names, Yes number of molecules types, No number of molecules of each type, No number of atoms of each molecule type, No coordinates, Yes charges Yes, if you use the right editconf option and masses on a .pdb file, correct? No. But, still, where could I then easilly get the box type and sizes, as well as simulation times? The .xtc file, the .edr file... Thanks a lot for all the help, and sorry for this punctual hurry here. Mark ___ gmx-users mailing listgmx-users@gromacs.org http://www.gromacs.org/mailman/listinfo/gmx-users Please search the archive at http://www.gromacs.org/search before posting! Please don't post (un)subscribe requests to the list. Use the www interface or send it to [EMAIL PROTECTED] Can't post? Read http://www.gromacs.org/mailing_lists/users.php
Re: [gmx-users] Reading XTC files from fortran90
Jones de Andrade wrote: Hi all. First: ok, lets stop the C X Fortran issue here: this kind of subject always look like a ready to become flamme war one. ;) Sorry for rising the subject a bit in the previous email. The point is: when this project started, almost 4 years ago, I already knew very well frotran90, and gromacs was not still in the sight for us, but fortran codes like mdynamix still were. So, at that point it was pointless to learn C. Three years later, at the end of the project, gromacs becomes all-day tool, and thousands of lines of code already written and optimized for fortran90 also are standard here. ;) That's the reason why it would be crazzyness to try to translate everything to C. So, the aim here is to just change the reading procedures, to read the information needed from gromacs. One way would be to rewrite everything on fortran, and that will be painfull. The other is to link to the objects and libraries in the proper way. If everything goes smoothly here, the next project I get in will include two years of implementing a few extra features inside gromacs code. Then, of course, learning C will be a must, and the first thing to be done. :) Fine :-) That's the sort of reason that you'd want to have to keep Fortran for the present problem. I can already do that for .xtc files. Seems that, unfortunatelly, the least odd way is to find a way to get the last pieces of information needed (atomic charges and masses) from the .tpr files. Other ways would include lots of file translations, or adaptions in the codes themselves which would lead to non-generality of it. Like I said, the masses are easy to infer from the atom names. You don't need to read them... you know anything starting with N has mass 14.011 (or whatever it is). Or you can read share/gromacs/top/atommass.dat to do the lookups. Sorry for not noticing the box shape in the .xtc file. I passed over that integer in the reading, dismissing it as only some kind of old feature useless now. Really sorry for that point. Which files should I look for in order to get the proper way of reading masses and atomic charges from a .tpr file? And which objects will need to be included in the linking stage? Is there any outsider program (fortran would be perfect, but I don't think it will be too common too) that works directly with the .tpr files? tpxio.c will have relevant functions. Or, follow the execution of mdrun until it reads the .tpr file and see how it handles these things. If you use the editconf suggestion to get charges, you don't need to read a .tpr file. Thanks a lot everybody for all the help, and sorry for taking so much of your time, Help is fine, but repeating myself is tiresome :-) Mark ___ gmx-users mailing listgmx-users@gromacs.org http://www.gromacs.org/mailman/listinfo/gmx-users Please search the archive at http://www.gromacs.org/search before posting! Please don't post (un)subscribe requests to the list. Use the www interface or send it to [EMAIL PROTECTED] Can't post? Read http://www.gromacs.org/mailing_lists/users.php
Re: [gmx-users] Reading XTC files from fortran90
Hi all. First of all, I would really want want to apologize if I somehow missunderstood any previous message from anyone, and made myself a bit annoying because of this fact. Second, I would like to say thanks again to eferybody for all help in these matters. On 10/1/07, Tsjerk Wassenaar [EMAIL PROTECTED] wrote: I'd say the easiest is to get into gmx_editconf and look at the option -grasp or -mead which put the atomic charge and vdw radius in the b-factor and occupancy field of a .pdb file when given a .tpr file. It's trivial to make it write charges and masses instead. Would also be a good first exercise for C/Gromacs :p Good idea! Even I don't choose to sitck with many file type conversions for the sake of the disk space, that specific program now seems to be a good start to learn explicitly how all the topology data is passed from one place to the other. And that would help me to learn how to read the topology files (since it seems to be only a matter now of reading atomic charges and masses). I'm not sure what you mean with shape and that integer. There's nothing much redundant in the .xtc file and there are no integers having to do with box shape. In fact, there is no shape... Well, by shape I meant cubic and truncated octahedral, for example. But the integer was my mistake (old dlpoly adaptions that just came to my mind, together with the first integer in the header of all xdrf files and that I could not find a proper meaning at time), thought for a moment that the first integer in the .xtc header mean the shape of the box. Just played a bit later with the manual and recorded that it's only the 9 box parameters that are important to define that within gromacs. Sorry for that mistake. So, I would have atom names, Yes From .gro or .pdb file in the standard way. number of molecules types, No number of molecules of each type, No number of atoms of each molecule type, No I think those three can become yes in a really easy way: some kind of state machine would easily do the trick and count them properly. coordinates, Yes From .xtc or .pdb file. charges Yes, if you use the right editconf option and masses on a .pdb file, correct? No. Only modifying editconf to print those in a adapted .pdb file, throw a file conversion and much disk space wasted. Or I can learn exactly from editconf how to read it from .tpr or .top (I guess it takes from these files the atomic charges when needed) and adapt into my codes. But, still, where could I then easilly get the box type and sizes, as well as simulation times? The .xtc file, the .edr file... So .xtc, since its already under control. ;) Fine :-) That's the sort of reason that you'd want to have to keep Fortran for the present problem. So we agree here. Three ways so: converting the trajectories to another format (ugly for my taste, and too much disk space needed), linking to a library (like done in reading .xtc files) or linking to an object. ;) Like I said, the masses are easy to infer from the atom names. You don't need to read them... you know anything starting with N has mass 14.011 (or whatever it is). Or you can read share/gromacs/top/atommass.dat to do the lookups. That is the kind of stuff I really don't like. For instance, all codes I written for all this time long aimed at being abolutelly general. So, if anything starting by N means nitrogen, how to deal when someone using the same program (even in the group) chooses to say that Nb is a different type of nitrogen while another says it is niobium? The program must be able to deal with such problems that arise (and believe, I've already got locked at this one once, and that's why I don't go such a way anymore), and the only way I see it doing so is directly reading the topologies, instead of knowing it in advance. ;) Sorry for not noticing the box shape in the .xtc file. I passed over that integer in the reading, dismissing it as only some kind of old feature useless now. Really sorry for that point. Which files should I look for in order to get the proper way of reading masses and atomic charges from a .tpr file? And which objects will need to be included in the linking stage? Is there any outsider program (fortran would be perfect, but I don't think it will be too common too) that works directly with the .tpr files? tpxio.c will have relevant functions. Or, follow the execution of mdrun until it reads the .tpr file and see how it handles these things. Thanks a lot! That final information will be really usefull for me to sort out how to properly link those routines in my codes. :) If you use the editconf suggestion to get charges, you don't need to read a .tpr file. But then or I get locked into the problem of generality of having too much disk usage. But that idea will be also usefull to learn how to read the masses and charges from .tpr files. Thanks a lot everybody for all the help, and sorry for taking so much
Re: [gmx-users] Reading XTC files from fortran90
Jones de Andrade wrote: From .gro or .pdb file in the standard way. number of molecules types, No number of molecules of each type, No number of atoms of each molecule type, No I think those three can become yes in a really easy way: some kind of state machine would easily do the trick and count them properly. Yes, but now you have to infer molecule changes from residue name changes, which is doable, but messy. coordinates, Yes From .xtc or .pdb file. charges Yes, if you use the right editconf option and masses on a .pdb file, correct? No. Only modifying editconf to print those in a adapted .pdb file, throw a file conversion and much disk space wasted. Or I can learn exactly from editconf how to read it from .tpr or .top (I guess it takes from these files the atomic charges when needed) and adapt into my codes. The point of editconf is to feed your system *one* pdb structure with atom names, charges and/or whatever - it would be easy to modify editconf to write masses in one column and charges in another. You rely on the user to then supply a .xtc file that corresponds to that .pdb, in the same way that most of the GROMACS analysis tools accept an .xtc and various subsets of the possible structure files. You don't convert your .xtc to a .pdb trajectory, because that would be immensely wasteful, and the static information of the names, masses and charges doesn't vary with time under the MD approximation. If you use the editconf suggestion to get charges, you don't need to read a .tpr file. But then or I get locked into the problem of generality of having too much disk usage. But that idea will be also usefull to learn how to read the masses and charges from .tpr files. Easier is to hack editconf as described above. Mark ___ gmx-users mailing listgmx-users@gromacs.org http://www.gromacs.org/mailman/listinfo/gmx-users Please search the archive at http://www.gromacs.org/search before posting! Please don't post (un)subscribe requests to the list. Use the www interface or send it to [EMAIL PROTECTED] Can't post? Read http://www.gromacs.org/mailing_lists/users.php
Re: [gmx-users] Reading XTC files from fortran90
Hi Jones, You probably only need to read the atoms and such from a .pdb or .gro file, unless you really want to have topology information (bondedness, molecule definitions, etc). Then you do have to read the topology file or a .tpr file directly. Don't recall having seen the fortran implementations for these... But maybe Bert knows better. Just one note... There's no shape in PBC. The lattice vectors are all there is, and these are stored in the .xtc file (as you already know). Cheers, Tsjerk On 9/30/07, Jones de Andrade [EMAIL PROTECTED] wrote: Hi Bert. Thanks a lot for the help. I've found the mistakes, one programming minor issue and a compiler configuration a bit of hidden in the config files. Old time mistakes just choose to arise at its worst moments. :) At the moment, I already can read the coordinates, number of atoms, box coordinates, step number, simulation time and precision. I guess this is all what is stored in the .xtc files themselves, am I right? I have so one extra question: I guess that extra information as atom names, number of molecules of type X, number of molecular types, box and pbc shape and atomic charges and masses are *not* stored in the .xtc files, am I right? Ok, the integer stuff (number of molecules of type X, number of molecular types, box and pbc) could be written in a special .ndx file, but that would, by all means, looks at least diselegant. If so, the only two possible ways of getting them is reading or the .top and .itp files correctly, or to read the .tpr file, correct? Now, is there any guideline available on how to deal with those files? Thanks a lot everybody for all the help. Sincerally yours, Jones On 9/29/07, Bert de Groot [EMAIL PROTECTED] wrote: On Sat, 29 Sep 2007, Jones de Andrade wrote: Hi Bert. Thank you for the prompt answer. Just did as instructed, but got the following: CruNumMac src/own/B/9/xtc 286% ifort xtciof.f90 test_xtc.f90 -L/home/johannes/src/own/B/9/xtc/xtc/ -lxrdf -lg2c IPO link: can not find -lxrdf well, apparently the path you specify with -L does not contain a libxdrf.a the way you called it the compiler expects it to be located here: /home/johannes/src/own/B/9/xtc/xtc/libxdrf.a also, IIRC, the xtcio stuff is already linked into the libxdrf.a, so no need to include it again. ifort: error: problem during multi-file optimization compilation (code 1) Looks better, in the sense that the number of error messages was reduced. But still doesn't accept to link to the xrdf library. Tried that with and without re-make of the library (strange fact that the SGI arch is to be used in linux) and also tried to say -llibxrdf instead of -lxrdf. Nothing worked. Have you or someone come across such an error before? Any clue of what can possibly be going wrong? Thanks a lot in advance... Sincerally yours, Jones On 9/29/07, Bert de Groot [EMAIL PROTECTED] wrote: Hi, try: download http://www.mpibpc.gwdg.de/groups/de_groot/xtc.tar.gz (and optionally issue a 'make' in the xtc directory after unpacking) in the linking stage, use something like ifort -blabla -lxdrf -L/wherever/xtc -lg2c your smalll test code looks OK apart from the fact that I don't know what happens if you readwrite from the same file. Bert Bert Bert de Groot, PhD Max Planck Institute for Biophysical Chemistry Computational biomolecular dynamics group Am Fassberg 11 37077 Goettingen, Germany tel: +49-551-2012308, fax: +49-551-2012302 http://www.mpibpc.gwdg.de/groups/de_groot ___ gmx-users mailing list gmx-users@gromacs.org http://www.gromacs.org/mailman/listinfo/gmx-users Please search the archive at http://www.gromacs.org/search before posting! Please don't post (un)subscribe requests to the list. Use the www interface or send it to [EMAIL PROTECTED] . Can't post? Read http://www.gromacs.org/mailing_lists/users.php ___ gmx-users mailing listgmx-users@gromacs.org http://www.gromacs.org/mailman/listinfo/gmx-users Please search the archive at http://www.gromacs.org/search before posting! Please don't post (un)subscribe requests to the list. Use the www interface or send it to [EMAIL PROTECTED] Can't post? Read http://www.gromacs.org/mailing_lists/users.php -- Tsjerk A. Wassenaar, Ph.D. Junior UD (post-doc) Biomolecular NMR, Bijvoet Center Utrecht University Padualaan 8 3584 CH Utrecht The Netherlands P: +31-30-2539931 F: +31-30-2537623 ___ gmx-users mailing listgmx-users@gromacs.org http://www.gromacs.org/mailman/listinfo/gmx-users Please search the archive at http://www.gromacs.org/search before posting! Please don't
Re: [gmx-users] Reading XTC files from fortran90
Hi Tsjerk. Thanks for the prompt answer. Good idea, I can take the atom names from a .gro file, nad, probably some strange algorithm can also get the number of atoms per molecule and number of molecules from there, am I right? That would help a lot. But: I'll still be unable to get atomic masses and point charges. No to mention the box shape. :( And I'll need a way to read these parameters too, unless I begin to write all them in a really awfull and strange index file. From what I can understand on the file logistics of gromacs, these information are only stored in the whole .top .itp .atp .rtp structure of files or, in a condensed form, in the .tpr file. Is that right? Thanks a lot for all the information and help! Sincerally yours, Jones On 9/30/07, Tsjerk Wassenaar [EMAIL PROTECTED] wrote: Hi Jones, You probably only need to read the atoms and such from a .pdb or .gro file, unless you really want to have topology information (bondedness, molecule definitions, etc). Then you do have to read the topology file or a .tpr file directly. Don't recall having seen the fortran implementations for these... But maybe Bert knows better. Just one note... There's no shape in PBC. The lattice vectors are all there is, and these are stored in the .xtc file (as you already know). Cheers, Tsjerk On 9/30/07, Jones de Andrade [EMAIL PROTECTED] wrote: Hi Bert. Thanks a lot for the help. I've found the mistakes, one programming minor issue and a compiler configuration a bit of hidden in the config files. Old time mistakes just choose to arise at its worst moments. :) At the moment, I already can read the coordinates, number of atoms, box coordinates, step number, simulation time and precision. I guess this is all what is stored in the .xtc files themselves, am I right? I have so one extra question: I guess that extra information as atom names, number of molecules of type X, number of molecular types, box and pbc shape and atomic charges and masses are *not* stored in the .xtc files, am I right? Ok, the integer stuff (number of molecules of type X, number of molecular types, box and pbc) could be written in a special .ndx file, but that would, by all means, looks at least diselegant. If so, the only two possible ways of getting them is reading or the .top and .itp files correctly, or to read the .tpr file, correct? Now, is there any guideline available on how to deal with those files? Thanks a lot everybody for all the help. Sincerally yours, Jones On 9/29/07, Bert de Groot [EMAIL PROTECTED] wrote: On Sat, 29 Sep 2007, Jones de Andrade wrote: Hi Bert. Thank you for the prompt answer. Just did as instructed, but got the following: CruNumMac src/own/B/9/xtc 286% ifort xtciof.f90 test_xtc.f90 -L/home/johannes/src/own/B/9/xtc/xtc/ -lxrdf -lg2c IPO link: can not find -lxrdf well, apparently the path you specify with -L does not contain a libxdrf.a the way you called it the compiler expects it to be located here: /home/johannes/src/own/B/9/xtc/xtc/libxdrf.a also, IIRC, the xtcio stuff is already linked into the libxdrf.a, so no need to include it again. ifort: error: problem during multi-file optimization compilation (code 1) Looks better, in the sense that the number of error messages was reduced. But still doesn't accept to link to the xrdf library. Tried that with and without re-make of the library (strange fact that the SGI arch is to be used in linux) and also tried to say -llibxrdf instead of -lxrdf. Nothing worked. Have you or someone come across such an error before? Any clue of what can possibly be going wrong? Thanks a lot in advance... Sincerally yours, Jones On 9/29/07, Bert de Groot [EMAIL PROTECTED] wrote: Hi, try: download http://www.mpibpc.gwdg.de/groups/de_groot/xtc.tar.gz (and optionally issue a 'make' in the xtc directory after unpacking) in the linking stage, use something like ifort -blabla -lxdrf -L/wherever/xtc -lg2c your smalll test code looks OK apart from the fact that I don't know what happens if you readwrite from the same file. Bert Bert Bert de Groot, PhD Max Planck Institute for Biophysical Chemistry Computational biomolecular dynamics group Am Fassberg 11 37077 Goettingen, Germany tel: +49-551-2012308, fax: +49-551-2012302 http://www.mpibpc.gwdg.de/groups/de_groot ___ gmx-users mailing list gmx-users@gromacs.org http://www.gromacs.org/mailman/listinfo/gmx-users Please search the archive at http://www.gromacs.org/search before posting! Please don't post (un)subscribe requests to
Re: [gmx-users] Reading XTC files from fortran90
Mark Abraham wrote: Jones de Andrade wrote: Hi Tsjerk. Thanks for the prompt answer. Good idea, I can take the atom names from a .gro file, nad, probably some strange algorithm can also get the number of atoms per molecule and number of molecules from there, am I right? That would help a lot. Simplest is to write a parser for a gmxdump of a .tpr file. This is what Perl was made for. If you're constrained to use Fortran for your analysis program, then developing an intermediate file format for it to parse might be worthwhile. But: I'll still be unable to get atomic masses and point charges. No to mention the box shape. :( And I'll need a way to read these parameters too, unless I begin to write all them in a really awfull and strange index file. From what I can understand on the file logistics of gromacs, these information are only stored in the whole .top .itp .atp .rtp structure of files or, in a condensed form, in the .tpr file. Is that right? Yeah. Writing something to do the lookups in the .*tp files which grompp does to form the .tpr is asking for trouble. Do the lookup in the .tpr file. The .edr file has box dimensions, of course. All this makes it sound like you should be dropping Fortran and writing in C or Perl! the box is in the xtc file, if you need the atom names a corresponding pdb or gro file should do the trick. don't try to read tpr files or gmxdumps of it, that's a complete waste of time. you can e.g. use editconf to dump a pdb file with charges in the b-factor fields. And indeed, using fortran will make your life quite miserable for these kind of things. -- David van der Spoel, Ph.D. Molec. Biophys. group, Dept. of Cell Molec. Biol., Uppsala University. Box 596, 75124 Uppsala, Sweden. Phone: +46184714205. Fax: +4618511755. [EMAIL PROTECTED] [EMAIL PROTECTED] http://folding.bmc.uu.se ___ gmx-users mailing listgmx-users@gromacs.org http://www.gromacs.org/mailman/listinfo/gmx-users Please search the archive at http://www.gromacs.org/search before posting! Please don't post (un)subscribe requests to the list. Use the www interface or send it to [EMAIL PROTECTED] Can't post? Read http://www.gromacs.org/mailing_lists/users.php
Re: [gmx-users] Reading XTC files from fortran90
David van der Spoel wrote: the box is in the xtc file, if you need the atom names a corresponding pdb or gro file should do the trick. don't try to read tpr files or gmxdumps of it, that's a complete waste of time. you can e.g. use editconf to dump a pdb file with charges in the b-factor fields. Yes, I now see that editconf has that ability, and Jones can fill in the masses himself, so he doesn't need to parse the .tpr at all. And indeed, using fortran will make your life quite miserable for these kind of things. :-) Mark ___ gmx-users mailing listgmx-users@gromacs.org http://www.gromacs.org/mailman/listinfo/gmx-users Please search the archive at http://www.gromacs.org/search before posting! Please don't post (un)subscribe requests to the list. Use the www interface or send it to [EMAIL PROTECTED] Can't post? Read http://www.gromacs.org/mailing_lists/users.php
Re: [gmx-users] Reading XTC files from fortran90
Hi all. Firstly, thanks a lot for all the help! ;) First, about reading the atom-names from a .gro file, I don't think it would so painfull. It would look like a bit to having a well programmed state machine in the code. About the language, yes, I admit that bor reads and writes, fortran is miserable. But for calculations (like the inner codes of a certain program? ;) only joking... :D ) if one chooses not to go by assembly, fortran is still the way, specially if you consider the extra simplicity of fortran90. Anyway, I would not even consider re-writing the reads in fortran, but would love to consider to link to a library (like in reading .xtc files) or gromacs compiled object just to read it correctly. ;) About the box, is it on the .xtc file? Could not get the proper variable of it, could someone give me some infor which is that? Finally: yes, I considered changing everything to a .pdb file, despite the file size and the fact that I completelly forgot it could have the charges on it. But even so I would still have problems with the atomic masses, am I right? Or can I put that on a standard pdb file with gromacs programs? Again, tahnks a lot for all help in advance! Sincerally yours, Jones On 9/30/07, David van der Spoel [EMAIL PROTECTED] wrote: Mark Abraham wrote: Jones de Andrade wrote: Hi Tsjerk. Thanks for the prompt answer. Good idea, I can take the atom names from a .gro file, nad, probably some strange algorithm can also get the number of atoms per molecule and number of molecules from there, am I right? That would help a lot. Simplest is to write a parser for a gmxdump of a .tpr file. This is what Perl was made for. If you're constrained to use Fortran for your analysis program, then developing an intermediate file format for it to parse might be worthwhile. But: I'll still be unable to get atomic masses and point charges. No to mention the box shape. :( And I'll need a way to read these parameters too, unless I begin to write all them in a really awfull and strange index file. From what I can understand on the file logistics of gromacs, these information are only stored in the whole .top .itp .atp .rtp structure of files or, in a condensed form, in the .tpr file. Is that right? Yeah. Writing something to do the lookups in the .*tp files which grompp does to form the .tpr is asking for trouble. Do the lookup in the .tpr file. The .edr file has box dimensions, of course. All this makes it sound like you should be dropping Fortran and writing in C or Perl! the box is in the xtc file, if you need the atom names a corresponding pdb or gro file should do the trick. don't try to read tpr files or gmxdumps of it, that's a complete waste of time. you can e.g. use editconf to dump a pdb file with charges in the b-factor fields. And indeed, using fortran will make your life quite miserable for these kind of things. -- David van der Spoel, Ph.D. Molec. Biophys. group, Dept. of Cell Molec. Biol., Uppsala University. Box 596, 75124 Uppsala, Sweden. Phone: +46184714205. Fax: +4618511755. [EMAIL PROTECTED][EMAIL PROTECTED] http://folding.bmc.uu.se ___ gmx-users mailing listgmx-users@gromacs.org http://www.gromacs.org/mailman/listinfo/gmx-users Please search the archive at http://www.gromacs.org/search before posting! Please don't post (un)subscribe requests to the list. Use the www interface or send it to [EMAIL PROTECTED] Can't post? Read http://www.gromacs.org/mailing_lists/users.php ___ gmx-users mailing listgmx-users@gromacs.org http://www.gromacs.org/mailman/listinfo/gmx-users Please search the archive at http://www.gromacs.org/search before posting! Please don't post (un)subscribe requests to the list. Use the www interface or send it to [EMAIL PROTECTED] Can't post? Read http://www.gromacs.org/mailing_lists/users.php
Re: [gmx-users] Reading XTC files from fortran90
Jones de Andrade wrote: Hi all. Firstly, thanks a lot for all the help! ;) First, about reading the atom-names from a .gro file, I don't think it would so painfull. It would look like a bit to having a well programmed state machine in the code. About the language, yes, I admit that bor reads and writes, fortran is miserable. But for calculations (like the inner codes of a certain program? ;) only joking... :D ) if one chooses not to go by assembly, fortran is still the way, specially if you consider the extra simplicity of fortran90. Anyway, I would not even consider re-writing the reads in fortran, but would love to consider to link to a library (like in reading .xtc files) or gromacs compiled object just to read it correctly. ;) So now you have to weigh up the amount of time your computer will be doing calculations (and then how much Fortran gains against the alternatives) vs the amount of your time you'll spend writing and debugging the I/O routines :-) Short of doing calculations whose cost is on the order of more MD simulations, I'd be amazed if Fortran comes out ahead. About the box, is it on the .xtc file? Could not get the proper variable of it, could someone give me some infor which is that? Check out the C I/O routines? Finally: yes, I considered changing everything to a .pdb file, despite the file size and the fact that I completelly forgot it could have the charges on it. But even so I would still have problems with the atomic masses, am I right? Or can I put that on a standard pdb file with gromacs programs? You've got atom names in the pdb, so look the masses up in your own code. Or post-process the pdb file to put masses in whichever of occupancy and B-factor doesn't have charges. Mark ___ gmx-users mailing listgmx-users@gromacs.org http://www.gromacs.org/mailman/listinfo/gmx-users Please search the archive at http://www.gromacs.org/search before posting! Please don't post (un)subscribe requests to the list. Use the www interface or send it to [EMAIL PROTECTED] Can't post? Read http://www.gromacs.org/mailing_lists/users.php
Re: [gmx-users] Reading XTC files from fortran90
Hi Mark. Wait a second, you mean that editconf would allow me to have both charges and masses on a .pdb file directly from command line (having other matters asking for my presence absolutelly now, so please forgive me if this message becomes somehow strange) So, I would have atom names, number of molecules types, number of molecules of each type, number of atoms of each molecule type, coordinates, charges and masses on a .pdb file, correct? But, still, where could I then easilly get the box type and sizes, as well as simulation times? Thanks a lot for all the help, and sorry for this punctual hurry here. Sincerally yours, Jones On 9/30/07, Mark Abraham [EMAIL PROTECTED] wrote: David van der Spoel wrote: the box is in the xtc file, if you need the atom names a corresponding pdb or gro file should do the trick. don't try to read tpr files or gmxdumps of it, that's a complete waste of time. you can e.g. use editconf to dump a pdb file with charges in the b-factor fields. Yes, I now see that editconf has that ability, and Jones can fill in the masses himself, so he doesn't need to parse the .tpr at all. And indeed, using fortran will make your life quite miserable for these kind of things. :-) Mark ___ gmx-users mailing listgmx-users@gromacs.org http://www.gromacs.org/mailman/listinfo/gmx-users Please search the archive at http://www.gromacs.org/search before posting! Please don't post (un)subscribe requests to the list. Use the www interface or send it to [EMAIL PROTECTED] Can't post? Read http://www.gromacs.org/mailing_lists/users.php ___ gmx-users mailing listgmx-users@gromacs.org http://www.gromacs.org/mailman/listinfo/gmx-users Please search the archive at http://www.gromacs.org/search before posting! Please don't post (un)subscribe requests to the list. Use the www interface or send it to [EMAIL PROTECTED] Can't post? Read http://www.gromacs.org/mailing_lists/users.php
Re: [gmx-users] Reading XTC files from fortran90
Hi Bert. Thank you for the prompt answer. Just did as instructed, but got the following: CruNumMac src/own/B/9/xtc 286% ifort xtciof.f90 test_xtc.f90 -L/home/johannes/src/own/B/9/xtc/xtc/ -lxrdf -lg2c IPO link: can not find -lxrdf ifort: error: problem during multi-file optimization compilation (code 1) Looks better, in the sense that the number of error messages was reduced. But still doesn't accept to link to the xrdf library. Tried that with and without re-make of the library (strange fact that the SGI arch is to be used in linux) and also tried to say -llibxrdf instead of -lxrdf. Nothing worked. Have you or someone come across such an error before? Any clue of what can possibly be going wrong? Thanks a lot in advance... Sincerally yours, Jones On 9/29/07, Bert de Groot [EMAIL PROTECTED] wrote: Hi, try: download http://www.mpibpc.gwdg.de/groups/de_groot/xtc.tar.gz (and optionally issue a 'make' in the xtc directory after unpacking) in the linking stage, use something like ifort -blabla -lxdrf -L/wherever/xtc -lg2c your smalll test code looks OK apart from the fact that I don't know what happens if you readwrite from the same file. Bert ___ gmx-users mailing listgmx-users@gromacs.org http://www.gromacs.org/mailman/listinfo/gmx-users Please search the archive at http://www.gromacs.org/search before posting! Please don't post (un)subscribe requests to the list. Use the www interface or send it to [EMAIL PROTECTED] Can't post? Read http://www.gromacs.org/mailing_lists/users.php
Re: [gmx-users] Reading XTC files from fortran90
On Sat, 29 Sep 2007, Jones de Andrade wrote: Hi Bert. Thank you for the prompt answer. Just did as instructed, but got the following: CruNumMac src/own/B/9/xtc 286% ifort xtciof.f90 test_xtc.f90 -L/home/johannes/src/own/B/9/xtc/xtc/ -lxrdf -lg2c IPO link: can not find -lxrdf well, apparently the path you specify with -L does not contain a libxdrf.a the way you called it the compiler expects it to be located here: /home/johannes/src/own/B/9/xtc/xtc/libxdrf.a also, IIRC, the xtcio stuff is already linked into the libxdrf.a, so no need to include it again. ifort: error: problem during multi-file optimization compilation (code 1) Looks better, in the sense that the number of error messages was reduced. But still doesn't accept to link to the xrdf library. Tried that with and without re-make of the library (strange fact that the SGI arch is to be used in linux) and also tried to say -llibxrdf instead of -lxrdf. Nothing worked. Have you or someone come across such an error before? Any clue of what can possibly be going wrong? Thanks a lot in advance... Sincerally yours, Jones On 9/29/07, Bert de Groot [EMAIL PROTECTED] wrote: Hi, try: download http://www.mpibpc.gwdg.de/groups/de_groot/xtc.tar.gz (and optionally issue a 'make' in the xtc directory after unpacking) in the linking stage, use something like ifort -blabla -lxdrf -L/wherever/xtc -lg2c your smalll test code looks OK apart from the fact that I don't know what happens if you readwrite from the same file. Bert Bert Bert de Groot, PhD Max Planck Institute for Biophysical Chemistry Computational biomolecular dynamics group Am Fassberg 11 37077 Goettingen, Germany tel: +49-551-2012308, fax: +49-551-2012302 http://www.mpibpc.gwdg.de/groups/de_groot ___ gmx-users mailing listgmx-users@gromacs.org http://www.gromacs.org/mailman/listinfo/gmx-users Please search the archive at http://www.gromacs.org/search before posting! Please don't post (un)subscribe requests to the list. Use the www interface or send it to [EMAIL PROTECTED] Can't post? Read http://www.gromacs.org/mailing_lists/users.php
Re: [gmx-users] Reading XTC files from fortran90
Hi Bert. Thanks a lot for the help. I've found the mistakes, one programming minor issue and a compiler configuration a bit of hidden in the config files. Old time mistakes just choose to arise at its worst moments. :) At the moment, I already can read the coordinates, number of atoms, box coordinates, step number, simulation time and precision. I guess this is all what is stored in the .xtc files themselves, am I right? I have so one extra question: I guess that extra information as atom names, number of molecules of type X, number of molecular types, box and pbc shape and atomic charges and masses are *not* stored in the .xtc files, am I right? Ok, the integer stuff (number of molecules of type X, number of molecular types, box and pbc) could be written in a special .ndx file, but that would, by all means, looks at least diselegant. If so, the only two possible ways of getting them is reading or the .top and .itp files correctly, or to read the .tpr file, correct? Now, is there any guideline available on how to deal with those files? Thanks a lot everybody for all the help. Sincerally yours, Jones On 9/29/07, Bert de Groot [EMAIL PROTECTED] wrote: On Sat, 29 Sep 2007, Jones de Andrade wrote: Hi Bert. Thank you for the prompt answer. Just did as instructed, but got the following: CruNumMac src/own/B/9/xtc 286% ifort xtciof.f90 test_xtc.f90 -L/home/johannes/src/own/B/9/xtc/xtc/ -lxrdf -lg2c IPO link: can not find -lxrdf well, apparently the path you specify with -L does not contain a libxdrf.a the way you called it the compiler expects it to be located here: /home/johannes/src/own/B/9/xtc/xtc/libxdrf.a also, IIRC, the xtcio stuff is already linked into the libxdrf.a, so no need to include it again. ifort: error: problem during multi-file optimization compilation (code 1) Looks better, in the sense that the number of error messages was reduced. But still doesn't accept to link to the xrdf library. Tried that with and without re-make of the library (strange fact that the SGI arch is to be used in linux) and also tried to say -llibxrdf instead of -lxrdf. Nothing worked. Have you or someone come across such an error before? Any clue of what can possibly be going wrong? Thanks a lot in advance... Sincerally yours, Jones On 9/29/07, Bert de Groot [EMAIL PROTECTED] wrote: Hi, try: download http://www.mpibpc.gwdg.de/groups/de_groot/xtc.tar.gz (and optionally issue a 'make' in the xtc directory after unpacking) in the linking stage, use something like ifort -blabla -lxdrf -L/wherever/xtc -lg2c your smalll test code looks OK apart from the fact that I don't know what happens if you readwrite from the same file. Bert Bert Bert de Groot, PhD Max Planck Institute for Biophysical Chemistry Computational biomolecular dynamics group Am Fassberg 11 37077 Goettingen, Germany tel: +49-551-2012308, fax: +49-551-2012302 http://www.mpibpc.gwdg.de/groups/de_groot ___ gmx-users mailing listgmx-users@gromacs.org http://www.gromacs.org/mailman/listinfo/gmx-users Please search the archive at http://www.gromacs.org/search before posting! Please don't post (un)subscribe requests to the list. Use the www interface or send it to [EMAIL PROTECTED] Can't post? Read http://www.gromacs.org/mailing_lists/users.php ___ gmx-users mailing listgmx-users@gromacs.org http://www.gromacs.org/mailman/listinfo/gmx-users Please search the archive at http://www.gromacs.org/search before posting! Please don't post (un)subscribe requests to the list. Use the www interface or send it to [EMAIL PROTECTED] Can't post? Read http://www.gromacs.org/mailing_lists/users.php
Re: [gmx-users] Reading XTC files from fortran90
Hi, try: download http://www.mpibpc.gwdg.de/groups/de_groot/xtc.tar.gz (and optionally issue a 'make' in the xtc directory after unpacking) in the linking stage, use something like ifort -blabla -lxdrf -L/wherever/xtc -lg2c your smalll test code looks OK apart from the fact that I don't know what happens if you readwrite from the same file. Bert On Sat, 29 Sep 2007, Jones de Andrade wrote: Hi all. I'm writing this beucase I'm having terrible problems in trying to deal with gromacs trajectory files with my own programs, originally written on fortran90. The codes themselves are enoughly big to make it out of question to try to translate them. What leads to some sort of mixed language compilation. My first guess was to try to use the guidelines from http://xray.bmc.uu.se/~spoel/md/online/xtc.html, but it absolutelly didn't worked. Could not find a compatible library. Tried to recompile gromacs in order to do so, but instead of a libsdrf.a library file, it created just common .lo and .o object files. Looked inside concoord program, as there are some mentions in the list to look at the code. Was a really good aprentize, could understand better the logics involved when dealing with .xtc files, BUT I could not use the library files that came with that. Somehow, it wasn't recognized as a valid file. My last try up to now was to try to link my test-read-program with the libxdrf object file. It yelded the following error: CruNumMac src/own/B/9/xtc 238% ifort libxdrf.lo xtciof.f90 test_xtc.f90 IPO: WARNING: no IR in object file libxdrf.lo; was the source file compiled with -ipo? IPO Error: unresolved : xdrfint_ Referenced in /tmp/ipo_ifortZNvZHh.o IPO Error: unresolved : xdrfclose_ Referenced in /tmp/ipo_ifortZNvZHh.o IPO Error: unresolved : xdrfopen_ Referenced in /tmp/ipo_ifortZNvZHh.o IPO Error: unresolved : xdrffloat_ Referenced in /tmp/ipo_ifortZNvZHh.o IPO Error: unresolved : xdrf3dfcoord_ Referenced in /tmp/ipo_ifortZNvZHh.o IPO: performing multi-file optimizations IPO: generating object file /tmp/ipo_ifortZNvZHh.o test_xtc.f90(17) : (col. 3) remark: LOOP WAS VECTORIZED. test_xtc.f90(25) : (col. 8) remark: LOOP WAS VECTORIZED. xtciof.f90(112) : (col. 3) remark: LOOP WAS VECTORIZED. libxdrf.lo: file not recognized: File format not recognized If I try to link to the .o objetc: CruNumMac src/own/B/9/xtc 240% ifort libxdrf.o xtciof.f90 test_xtc.f90 IPO: WARNING: no IR in object file libxdrf.o; was the source file compiled with -ipo? IPO Error: unresolved : xdrfint_ Referenced in /tmp/ipo_ifort2f5KMw.o IPO Error: unresolved : xdrfclose_ Referenced in /tmp/ipo_ifort2f5KMw.o IPO Error: unresolved : xdrfopen_ Referenced in /tmp/ipo_ifort2f5KMw.o IPO Error: unresolved : xdrffloat_ Referenced in /tmp/ipo_ifort2f5KMw.o IPO Error: unresolved : xdrf3dfcoord_ Referenced in /tmp/ipo_ifort2f5KMw.o IPO Error: unresolved : ffclose Referenced in libxdrf.o IPO: performing multi-file optimizations IPO: generating object file /tmp/ipo_ifort2f5KMw.o test_xtc.f90(17) : (col. 3) remark: LOOP WAS VECTORIZED. test_xtc.f90(25) : (col. 8) remark: LOOP WAS VECTORIZED. xtciof.f90(112) : (col. 3) remark: LOOP WAS VECTORIZED. /tmp/ipo_ifort2f5KMw.o(.text+0x1e1): In function `MAIN__': /tmp/ipo_ifort2f5KMw.f: undefined reference to `xdrfopen_' /tmp/ipo_ifort2f5KMw.o(.text+0x4ea):/tmp/ipo_ifort2f5KMw.f: undefined reference to `xdrfclose_' /tmp/ipo_ifort2f5KMw.o(.text+0x517):/tmp/ipo_ifort2f5KMw.f: undefined reference to `xdrfopen_' /tmp/ipo_ifort2f5KMw.o(.text+0x92b): In function `xtciof_mp_xtcheader_': /tmp/ipo_ifort2f5KMw.f: undefined reference to `xdrfint_' /tmp/ipo_ifort2f5KMw.o(.text+0x941):/tmp/ipo_ifort2f5KMw.f: undefined reference to `xdrfint_' /tmp/ipo_ifort2f5KMw.o(.text+0x95e):/tmp/ipo_ifort2f5KMw.f: undefined reference to `xdrfint_' /tmp/ipo_ifort2f5KMw.o(.text+0x97b):/tmp/ipo_ifort2f5KMw.f: undefined reference to `xdrffloat_' /tmp/ipo_ifort2f5KMw.o(.text+0xb8e): In function `xtciof_mp_xtcio_': /tmp/ipo_ifort2f5KMw.f: undefined reference to `xdrffloat_' /tmp/ipo_ifort2f5KMw.o(.text+0xbde):/tmp/ipo_ifort2f5KMw.f: undefined reference to `xdrffloat_' /tmp/ipo_ifort2f5KMw.o(.text+0xc2e):/tmp/ipo_ifort2f5KMw.f: undefined reference to `xdrffloat_' /tmp/ipo_ifort2f5KMw.o(.text+0xc7e):/tmp/ipo_ifort2f5KMw.f: undefined reference to `xdrffloat_' /tmp/ipo_ifort2f5KMw.o(.text+0xcce):/tmp/ipo_ifort2f5KMw.f: more undefined references to `xdrffloat_' follow /tmp/ipo_ifort2f5KMw.o(.text+0xe6b): In function `xtciof_mp_xtcio_': /tmp/ipo_ifort2f5KMw.f: undefined reference to `xdrf3dfcoord_' /tmp/ipo_ifort2f5KMw.o(.text+0xfc6): In function `xtciof_mp_xtcopen_': /tmp/ipo_ifort2f5KMw.f: undefined reference to `xdrfopen_' /tmp/ipo_ifort2f5KMw.o(.text+0x1292):/tmp/ipo_ifort2f5KMw.f: undefined reference to `xdrfclose_'