2 * This file is part of the GROMACS molecular simulation package.
4 * Copyright (c) 1991-2000, University of Groningen, The Netherlands.
5 * Copyright (c) 2001-2004, The GROMACS development team.
6 * Copyright (c) 2013,2014,2015, by the GROMACS development team, led by
7 * Mark Abraham, David van der Spoel, Berk Hess, and Erik Lindahl,
8 * and including many others, as listed in the AUTHORS file in the
9 * top-level source directory and at http://www.gromacs.org.
11 * GROMACS is free software; you can redistribute it and/or
12 * modify it under the terms of the GNU Lesser General Public License
13 * as published by the Free Software Foundation; either version 2.1
14 * of the License, or (at your option) any later version.
16 * GROMACS is distributed in the hope that it will be useful,
17 * but WITHOUT ANY WARRANTY; without even the implied warranty of
18 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
19 * Lesser General Public License for more details.
21 * You should have received a copy of the GNU Lesser General Public
22 * License along with GROMACS; if not, see
23 * http://www.gnu.org/licenses, or write to the Free Software Foundation,
24 * Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA.
26 * If you want to redistribute modifications to GROMACS, please
27 * consider that scientific software is very special. Version
28 * control is crucial - bugs must be traceable. We will be happy to
29 * consider code for inclusion in the official distribution, but
30 * derived work must not be called official GROMACS. Details are found
31 * in the README & COPYING files - if they are missing, get the
32 * official version at http://www.gromacs.org.
34 * To help us fund GROMACS development, we humbly ask that you cite
35 * the research papers on the package. Check out http://www.gromacs.org.
42 #include "gromacs/commandline/pargs.h"
43 #include "gromacs/fileio/mtxio.h"
44 #include "gromacs/fileio/xvgr.h"
45 #include "gromacs/gmxana/eigio.h"
46 #include "gromacs/gmxana/gmx_ana.h"
47 #include "gromacs/gmxana/gstat.h"
48 #include "gromacs/legacyheaders/copyrite.h"
49 #include "gromacs/legacyheaders/macros.h"
50 #include "gromacs/legacyheaders/txtdump.h"
51 #include "gromacs/legacyheaders/typedefs.h"
52 #include "gromacs/linearalgebra/eigensolver.h"
53 #include "gromacs/linearalgebra/sparsematrix.h"
54 #include "gromacs/math/units.h"
55 #include "gromacs/math/vec.h"
56 #include "gromacs/topology/mtop_util.h"
57 #include "gromacs/utility/futil.h"
58 #include "gromacs/utility/gmxassert.h"
59 #include "gromacs/utility/smalloc.h"
61 static double cv_corr(double nu, double T)
63 double x = PLANCK*nu/(BOLTZ*T);
64 double ex = std::exp(x);
72 return BOLTZ*KILO*(ex*sqr(x)/sqr(ex-1) - 1);
76 static double u_corr(double nu, double T)
78 double x = PLANCK*nu/(BOLTZ*T);
79 double ex = std::exp(x);
87 return BOLTZ*T*(0.5*x - 1 + x/(ex-1));
91 static int get_nharm_mt(gmx_moltype_t *mt)
93 static int harm_func[] = { F_BONDS };
97 for (i = 0; (i < asize(harm_func)); i++)
100 nh += mt->ilist[ft].nr/(interaction_function[ft].nratoms+1);
105 static int get_nvsite_mt(gmx_moltype_t *mt)
107 static int vs_func[] = {
108 F_VSITE2, F_VSITE3, F_VSITE3FD, F_VSITE3FAD,
109 F_VSITE3OUT, F_VSITE4FD, F_VSITE4FDN, F_VSITEN
114 for (i = 0; (i < asize(vs_func)); i++)
117 nh += mt->ilist[ft].nr/(interaction_function[ft].nratoms+1);
122 static int get_nharm(gmx_mtop_t *mtop, int *nvsites)
128 for (j = 0; (j < mtop->nmolblock); j++)
130 mt = mtop->molblock[j].type;
131 nh += mtop->molblock[j].nmol * get_nharm_mt(&(mtop->moltype[mt]));
132 nv += mtop->molblock[j].nmol * get_nvsite_mt(&(mtop->moltype[mt]));
139 nma_full_hessian(real * hess,
152 natoms = top->atoms.nr;
154 /* divide elements hess[i][j] by sqrt(mas[i])*sqrt(mas[j]) when required */
158 for (i = 0; (i < natoms); i++)
160 for (j = 0; (j < DIM); j++)
162 for (k = 0; (k < natoms); k++)
164 mass_fac = gmx_invsqrt(top->atoms.atom[i].m*top->atoms.atom[k].m);
165 for (l = 0; (l < DIM); l++)
167 hess[(i*DIM+j)*ndim+k*DIM+l] *= mass_fac;
174 /* call diagonalization routine. */
176 fprintf(stderr, "\nDiagonalizing to find vectors %d through %d...\n", begin, end);
179 eigensolver(hess, ndim, begin-1, end-1, eigenvalues, eigenvectors);
181 /* And scale the output eigenvectors */
182 if (bM && eigenvectors != NULL)
184 for (i = 0; i < (end-begin+1); i++)
186 for (j = 0; j < natoms; j++)
188 mass_fac = gmx_invsqrt(top->atoms.atom[j].m);
189 for (k = 0; (k < DIM); k++)
191 eigenvectors[i*ndim+j*DIM+k] *= mass_fac;
201 nma_sparse_hessian(gmx_sparsematrix_t * sparse_hessian,
215 natoms = top->atoms.nr;
218 /* Cannot check symmetry since we only store half matrix */
219 /* divide elements hess[i][j] by sqrt(mas[i])*sqrt(mas[j]) when required */
221 GMX_RELEASE_ASSERT(sparse_hessian != NULL, "NULL matrix pointer provided to nma_sparse_hessian");
225 for (iatom = 0; (iatom < natoms); iatom++)
227 for (j = 0; (j < DIM); j++)
230 for (k = 0; k < sparse_hessian->ndata[row]; k++)
232 col = sparse_hessian->data[row][k].col;
234 mass_fac = gmx_invsqrt(top->atoms.atom[iatom].m*top->atoms.atom[katom].m);
235 sparse_hessian->data[row][k].value *= mass_fac;
240 fprintf(stderr, "\nDiagonalizing to find eigenvectors 1 through %d...\n", neig);
243 sparse_eigensolver(sparse_hessian, neig, eigenvalues, eigenvectors, 10000000);
245 /* Scale output eigenvectors */
246 if (bM && eigenvectors != NULL)
248 for (i = 0; i < neig; i++)
250 for (j = 0; j < natoms; j++)
252 mass_fac = gmx_invsqrt(top->atoms.atom[j].m);
253 for (k = 0; (k < DIM); k++)
255 eigenvectors[i*ndim+j*DIM+k] *= mass_fac;
264 int gmx_nmeig(int argc, char *argv[])
266 const char *desc[] = {
267 "[THISMODULE] calculates the eigenvectors/values of a (Hessian) matrix,",
268 "which can be calculated with [gmx-mdrun].",
269 "The eigenvectors are written to a trajectory file ([TT]-v[tt]).",
270 "The structure is written first with t=0. The eigenvectors",
271 "are written as frames with the eigenvector number as timestamp.",
272 "The eigenvectors can be analyzed with [gmx-anaeig].",
273 "An ensemble of structures can be generated from the eigenvectors with",
274 "[gmx-nmens]. When mass weighting is used, the generated eigenvectors",
275 "will be scaled back to plain Cartesian coordinates before generating the",
276 "output. In this case, they will no longer be exactly orthogonal in the",
277 "standard Cartesian norm, but in the mass-weighted norm they would be.[PAR]",
278 "This program can be optionally used to compute quantum corrections to heat capacity",
279 "and enthalpy by providing an extra file argument [TT]-qcorr[tt]. See the GROMACS",
280 "manual, Chapter 1, for details. The result includes subtracting a harmonic",
281 "degree of freedom at the given temperature.",
282 "The total correction is printed on the terminal screen.",
283 "The recommended way of getting the corrections out is:[PAR]",
284 "[TT]gmx nmeig -s topol.tpr -f nm.mtx -first 7 -last 10000 -T 300 -qc [-constr][tt][PAR]",
285 "The [TT]-constr[tt] option should be used when bond constraints were used during the",
286 "simulation [BB]for all the covalent bonds[bb]. If this is not the case, ",
287 "you need to analyze the [TT]quant_corr.xvg[tt] file yourself.[PAR]",
288 "To make things more flexible, the program can also take virtual sites into account",
289 "when computing quantum corrections. When selecting [TT]-constr[tt] and",
290 "[TT]-qc[tt], the [TT]-begin[tt] and [TT]-end[tt] options will be set automatically as well.",
291 "Again, if you think you know it better, please check the [TT]eigenfreq.xvg[tt]",
295 static gmx_bool bM = TRUE, bCons = FALSE;
296 static int begin = 1, end = 50, maxspec = 4000;
297 static real T = 298.15, width = 1;
300 { "-m", FALSE, etBOOL, {&bM},
301 "Divide elements of Hessian by product of sqrt(mass) of involved "
302 "atoms prior to diagonalization. This should be used for 'Normal Modes' "
304 { "-first", FALSE, etINT, {&begin},
305 "First eigenvector to write away" },
306 { "-last", FALSE, etINT, {&end},
307 "Last eigenvector to write away" },
308 { "-maxspec", FALSE, etINT, {&maxspec},
309 "Highest frequency (1/cm) to consider in the spectrum" },
310 { "-T", FALSE, etREAL, {&T},
311 "Temperature for computing quantum heat capacity and enthalpy when using normal mode calculations to correct classical simulations" },
312 { "-constr", FALSE, etBOOL, {&bCons},
313 "If constraints were used in the simulation but not in the normal mode analysis (this is the recommended way of doing it) you will need to set this for computing the quantum corrections." },
314 { "-width", FALSE, etREAL, {&width},
315 "Width (sigma) of the gaussian peaks (1/cm) when generating a spectrum" }
317 FILE *out, *qc, *spec;
324 real qcvtot, qutot, qcv, qu;
325 int natoms, ndim, nrow, ncol, nharm, nvsite;
329 int version, generation;
330 real value, omega, nu;
331 real factor_gmx_to_omega2;
332 real factor_omega_to_wavenumber;
333 real *spectrum = NULL;
336 const char *qcleg[] = {
337 "Heat Capacity cV (J/mol K)",
338 "Enthalpy H (kJ/mol)"
340 real * full_hessian = NULL;
341 gmx_sparsematrix_t * sparse_hessian = NULL;
344 { efMTX, "-f", "hessian", ffREAD },
345 { efTPR, NULL, NULL, ffREAD },
346 { efXVG, "-of", "eigenfreq", ffWRITE },
347 { efXVG, "-ol", "eigenval", ffWRITE },
348 { efXVG, "-os", "spectrum", ffOPTWR },
349 { efXVG, "-qc", "quant_corr", ffOPTWR },
350 { efTRN, "-v", "eigenvec", ffWRITE }
352 #define NFILE asize(fnm)
354 if (!parse_common_args(&argc, argv, 0,
355 NFILE, fnm, asize(pa), pa, asize(desc), desc, 0, NULL, &oenv))
360 /* Read tpr file for volume and number of harmonic terms */
361 read_tpxheader(ftp2fn(efTPR, NFILE, fnm), &tpx, TRUE, &version, &generation);
362 snew(top_x, tpx.natoms);
364 read_tpx(ftp2fn(efTPR, NFILE, fnm), NULL, box, &natoms,
365 top_x, NULL, NULL, &mtop);
368 nharm = get_nharm(&mtop, &nvsite);
375 top = gmx_mtop_t_to_t_topology(&mtop);
380 if (opt2bSet("-qc", NFILE, fnm))
382 begin = 7+DIM*nvsite;
393 printf("Using begin = %d and end = %d\n", begin, end);
395 /*open Hessian matrix */
396 gmx_mtxio_read(ftp2fn(efMTX, NFILE, fnm), &nrow, &ncol, &full_hessian, &sparse_hessian);
398 /* Memory for eigenvalues and eigenvectors (begin..end) */
399 snew(eigenvalues, nrow);
400 snew(eigenvectors, nrow*(end-begin+1));
402 /* If the Hessian is in sparse format we can calculate max (ndim-1) eigenvectors,
403 * and they must start at the first one. If this is not valid we convert to full matrix
404 * storage, but warn the user that we might run out of memory...
406 if ((sparse_hessian != NULL) && (begin != 1 || end == ndim))
410 fprintf(stderr, "Cannot use sparse Hessian with first eigenvector != 1.\n");
412 else if (end == ndim)
414 fprintf(stderr, "Cannot use sparse Hessian to calculate all eigenvectors.\n");
417 fprintf(stderr, "Will try to allocate memory and convert to full matrix representation...\n");
419 snew(full_hessian, nrow*ncol);
420 for (i = 0; i < nrow*ncol; i++)
425 for (i = 0; i < sparse_hessian->nrow; i++)
427 for (j = 0; j < sparse_hessian->ndata[i]; j++)
429 k = sparse_hessian->data[i][j].col;
430 value = sparse_hessian->data[i][j].value;
431 full_hessian[i*ndim+k] = value;
432 full_hessian[k*ndim+i] = value;
435 gmx_sparsematrix_destroy(sparse_hessian);
436 sparse_hessian = NULL;
437 fprintf(stderr, "Converted sparse to full matrix storage.\n");
440 if (full_hessian != NULL)
442 /* Using full matrix storage */
443 nma_full_hessian(full_hessian, nrow, bM, &top, begin, end,
444 eigenvalues, eigenvectors);
448 /* Sparse memory storage, allocate memory for eigenvectors */
449 snew(eigenvectors, ncol*end);
450 nma_sparse_hessian(sparse_hessian, bM, &top, end, eigenvalues, eigenvectors);
453 /* check the output, first 6 eigenvalues should be reasonably small */
455 for (i = begin-1; (i < 6); i++)
457 if (std::abs(eigenvalues[i]) > 1.0e-3)
464 fprintf(stderr, "\nOne of the lowest 6 eigenvalues has a non-zero value.\n");
465 fprintf(stderr, "This could mean that the reference structure was not\n");
466 fprintf(stderr, "properly energy minimized.\n");
469 /* now write the output */
470 fprintf (stderr, "Writing eigenvalues...\n");
471 out = xvgropen(opt2fn("-ol", NFILE, fnm),
472 "Eigenvalues", "Eigenvalue index", "Eigenvalue [Gromacs units]",
474 if (output_env_get_print_xvgr_codes(oenv))
478 fprintf(out, "@ subtitle \"mass weighted\"\n");
482 fprintf(out, "@ subtitle \"not mass weighted\"\n");
486 for (i = 0; i <= (end-begin); i++)
488 fprintf (out, "%6d %15g\n", begin+i, eigenvalues[i]);
493 if (opt2bSet("-qc", NFILE, fnm))
495 qc = xvgropen(opt2fn("-qc", NFILE, fnm), "Quantum Corrections", "Eigenvector index", "", oenv);
496 xvgr_legend(qc, asize(qcleg), qcleg, oenv);
503 printf("Writing eigenfrequencies - negative eigenvalues will be set to zero.\n");
505 out = xvgropen(opt2fn("-of", NFILE, fnm),
506 "Eigenfrequencies", "Eigenvector index", "Wavenumber [cm\\S-1\\N]",
508 if (output_env_get_print_xvgr_codes(oenv))
512 fprintf(out, "@ subtitle \"mass weighted\"\n");
516 fprintf(out, "@ subtitle \"not mass weighted\"\n");
521 if (opt2bSet("-os", NFILE, fnm) && (maxspec > 0))
523 snew(spectrum, maxspec);
524 spec = xvgropen(opt2fn("-os", NFILE, fnm),
525 "Vibrational spectrum based on harmonic approximation",
526 "\\f{12}w\\f{4} (cm\\S-1\\N)",
527 "Intensity [Gromacs units]",
529 for (i = 0; (i < maxspec); i++)
535 /* Gromacs units are kJ/(mol*nm*nm*amu),
536 * where amu is the atomic mass unit.
538 * For the eigenfrequencies we want to convert this to spectroscopic absorption
539 * wavenumbers given in cm^(-1), which is the frequency divided by the speed of
540 * light. Do this by first converting to omega^2 (units 1/s), take the square
541 * root, and finally divide by the speed of light (nm/ps in gromacs).
543 factor_gmx_to_omega2 = 1.0E21/(AVOGADRO*AMU);
544 factor_omega_to_wavenumber = 1.0E-5/(2.0*M_PI*SPEED_OF_LIGHT);
546 for (i = begin; (i <= end); i++)
548 value = eigenvalues[i-begin];
553 omega = std::sqrt(value*factor_gmx_to_omega2);
554 nu = 1e-12*omega/(2*M_PI);
555 value = omega*factor_omega_to_wavenumber;
556 fprintf (out, "%6d %15g\n", i, value);
559 wfac = eigenvalues[i-begin]/(width*std::sqrt(2*M_PI));
560 for (j = 0; (j < maxspec); j++)
562 spectrum[j] += wfac*std::exp(-sqr(j-value)/(2*sqr(width)));
567 qcv = cv_corr(nu, T);
574 fprintf (qc, "%6d %15g %15g\n", i, qcv, qu);
582 for (j = 0; (j < maxspec); j++)
584 fprintf(spec, "%10g %10g\n", 1.0*j, spectrum[j]);
590 printf("Quantum corrections for harmonic degrees of freedom\n");
591 printf("Use appropriate -first and -last options to get reliable results.\n");
592 printf("There were %d constraints and %d vsites in the simulation\n",
594 printf("Total correction to cV = %g J/mol K\n", qcvtot);
595 printf("Total correction to H = %g kJ/mol\n", qutot);
597 please_cite(stdout, "Caleman2011b");
599 /* Writing eigenvectors. Note that if mass scaling was used, the eigenvectors
600 * were scaled back from mass weighted cartesian to plain cartesian in the
601 * nma_full_hessian() or nma_sparse_hessian() routines. Mass scaled vectors
602 * will not be strictly orthogonal in plain cartesian scalar products.
604 write_eigenvectors(opt2fn("-v", NFILE, fnm), natoms, eigenvectors, FALSE, begin, end,
605 eWXR_NO, NULL, FALSE, top_x, bM, eigenvalues);