int gmx_mdmat(int argc, char *argv[])
{
const char *desc[] = {
"[TT]g_mdmat[tt] makes distance matrices consisting of the smallest distance",
"between residue pairs. With [TT]-frames[tt], these distance matrices can be",
"stored in order to see differences in tertiary structure as a",
"function of time. If you choose your options unwisely, this may generate",
"a large output file. By default, only an averaged matrix over the whole",
"trajectory is output.",
"Also a count of the number of different atomic contacts between",
"residues over the whole trajectory can be made.",
"The output can be processed with [TT]xpm2ps[tt] to make a PostScript (tm) plot."
};
static real truncate = 1.5;
static gmx_bool bAtom = FALSE;
static int nlevels = 40;
t_pargs pa[] = {
{ "-t", FALSE, etREAL, {&truncate},
"trunc distance" },
{ "-nlevels", FALSE, etINT, {&nlevels},
"Discretize distance in this number of levels" }
};
t_filenm fnm[] = {
{ efTRX, "-f", NULL, ffREAD },
{ efTPS, NULL, NULL, ffREAD },
{ efNDX, NULL, NULL, ffOPTRD },
{ efXPM, "-mean", "dm", ffWRITE },
{ efXPM, "-frames", "dmf", ffOPTWR },
{ efXVG, "-no", "num", ffOPTWR },
};
#define NFILE asize(fnm)
FILE *out = NULL, *fp;
t_topology top;
int ePBC;
t_atoms useatoms;
int isize;
atom_id *index;
char *grpname;
int *rndx, *natm, prevres, newres;
int i, j, nres, natoms, nframes, it, trxnat;
t_trxstatus *status;
int nr0;
gmx_bool bCalcN, bFrames;
real t, ratio;
char title[256], label[234];
t_rgb rlo, rhi;
rvec *x;
real **mdmat, *resnr, **totmdmat;
int **nmat, **totnmat;
real *mean_n;
int *tot_n;
matrix box;
output_env_t oenv;
gmx_rmpbc_t gpbc = NULL;
CopyRight(stderr, argv[0]);
parse_common_args(&argc, argv, PCA_CAN_TIME | PCA_BE_NICE, NFILE, fnm,
asize(pa), pa, asize(desc), desc, 0, NULL, &oenv);
fprintf(stderr, "Will truncate at %f nm\n", truncate);
bCalcN = opt2bSet("-no", NFILE, fnm);
bFrames = opt2bSet("-frames", NFILE, fnm);
if (bCalcN)
{
fprintf(stderr, "Will calculate number of different contacts\n");
}
read_tps_conf(ftp2fn(efTPS, NFILE, fnm), title, &top, &ePBC, &x, NULL, box, FALSE);
fprintf(stderr, "Select group for analysis\n");
get_index(&top.atoms, ftp2fn_null(efNDX, NFILE, fnm), 1, &isize, &index, &grpname);
natoms = isize;
snew(useatoms.atom, natoms);
snew(useatoms.atomname, natoms);
useatoms.nres = 0;
snew(useatoms.resinfo, natoms);
prevres = top.atoms.atom[index[0]].resind;
newres = 0;
for (i = 0; (i < isize); i++)
{
int ii = index[i];
useatoms.atomname[i] = top.atoms.atomname[ii];
if (top.atoms.atom[ii].resind != prevres)
{
prevres = top.atoms.atom[ii].resind;
newres++;
useatoms.resinfo[i] = top.atoms.resinfo[prevres];
if (debug)
{
fprintf(debug, "New residue: atom %5s %5s %6d, index entry %5d, newres %5d\n",
*(top.atoms.resinfo[top.atoms.atom[ii].resind].name),
*(top.atoms.atomname[ii]),
ii, i, newres);
}
}
useatoms.atom[i].resind = newres;
}
useatoms.nres = newres+1;
useatoms.nr = isize;
rndx = res_ndx(&(useatoms));
natm = res_natm(&(useatoms));
nres = useatoms.nres;
fprintf(stderr, "There are %d residues with %d atoms\n", nres, natoms);
snew(resnr, nres);
snew(mdmat, nres);
snew(nmat, nres);
snew(totnmat, nres);
snew(mean_n, nres);
snew(tot_n, nres);
for (i = 0; (i < nres); i++)
{
snew(mdmat[i], nres);
snew(nmat[i], natoms);
snew(totnmat[i], natoms);
resnr[i] = i+1;
}
snew(totmdmat, nres);
for (i = 0; (i < nres); i++)
{
snew(totmdmat[i], nres);
}
trxnat = read_first_x(oenv, &status, ftp2fn(efTRX, NFILE, fnm), &t, &x, box);
nframes = 0;
rlo.r = 1.0, rlo.g = 1.0, rlo.b = 1.0;
rhi.r = 0.0, rhi.g = 0.0, rhi.b = 0.0;
gpbc = gmx_rmpbc_init(&top.idef, ePBC, trxnat, box);
if (bFrames)
{
out = opt2FILE("-frames", NFILE, fnm, "w");
}
do
{
gmx_rmpbc(gpbc, trxnat, box, x);
nframes++;
calc_mat(nres, natoms, rndx, x, index, truncate, mdmat, nmat, ePBC, box);
for (i = 0; (i < nres); i++)
{
for (j = 0; (j < natoms); j++)
{
if (nmat[i][j])
{
totnmat[i][j]++;
}
}
}
for (i = 0; (i < nres); i++)
{
for (j = 0; (j < nres); j++)
{
totmdmat[i][j] += mdmat[i][j];
}
}
if (bFrames)
{
sprintf(label, "t=%.0f ps", t);
write_xpm(out, 0, label, "Distance (nm)", "Residue Index", "Residue Index",
nres, nres, resnr, resnr, mdmat, 0, truncate, rlo, rhi, &nlevels);
}
}
while (read_next_x(oenv, status, &t, trxnat, x, box));
fprintf(stderr, "\n");
close_trj(status);
gmx_rmpbc_done(gpbc);
if (bFrames)
{
ffclose(out);
}
fprintf(stderr, "Processed %d frames\n", nframes);
for (i = 0; (i < nres); i++)
{
for (j = 0; (j < nres); j++)
{
totmdmat[i][j] /= nframes;
}
}
write_xpm(opt2FILE("-mean", NFILE, fnm, "w"), 0, "Mean smallest distance",
"Distance (nm)", "Residue Index", "Residue Index",
nres, nres, resnr, resnr, totmdmat, 0, truncate, rlo, rhi, &nlevels);
if (bCalcN)
{
tot_nmat(nres, natoms, nframes, totnmat, tot_n, mean_n);
fp = xvgropen(ftp2fn(efXVG, NFILE, fnm),
"Increase in number of contacts", "Residue", "Ratio", oenv);
fprintf(fp, "@ legend on\n");
fprintf(fp, "@ legend box on\n");
fprintf(fp, "@ legend loctype view\n");
fprintf(fp, "@ legend 0.75, 0.8\n");
fprintf(fp, "@ legend string 0 \"Total/mean\"\n");
fprintf(fp, "@ legend string 1 \"Total\"\n");
fprintf(fp, "@ legend string 2 \"Mean\"\n");
fprintf(fp, "@ legend string 3 \"# atoms\"\n");
fprintf(fp, "@ legend string 4 \"Mean/# atoms\"\n");
fprintf(fp, "#%3s %8s %3s %8s %3s %8s\n",
"res", "ratio", "tot", "mean", "natm", "mean/atm");
for (i = 0; (i < nres); i++)
{
if (mean_n[i] == 0)
{
ratio = 1;
}
else
{
ratio = tot_n[i]/mean_n[i];
}
fprintf(fp, "%3d %8.3f %3d %8.3f %3d %8.3f\n",
i+1, ratio, tot_n[i], mean_n[i], natm[i], mean_n[i]/natm[i]);
}
ffclose(fp);
}
thanx(stderr);
return 0;
}
int gmx_do_dssp(int argc, char *argv[])
{
const char *desc[] = {
"[THISMODULE] ",
"reads a trajectory file and computes the secondary structure for",
"each time frame ",
"calling the dssp program. If you do not have the dssp program,",
"get it from http://swift.cmbi.ru.nl/gv/dssp. [THISMODULE] assumes ",
"that the dssp executable is located in ",
"[TT]/usr/local/bin/dssp[tt]. If this is not the case, then you should",
"set an environment variable [TT]DSSP[tt] pointing to the dssp",
"executable, e.g.: [PAR]",
"[TT]setenv DSSP /opt/dssp/bin/dssp[tt][PAR]",
"Since version 2.0.0, dssp is invoked with a syntax that differs",
"from earlier versions. If you have an older version of dssp,",
"use the [TT]-ver[tt] option to direct do_dssp to use the older syntax.",
"By default, do_dssp uses the syntax introduced with version 2.0.0.",
"Even newer versions (which at the time of writing are not yet released)",
"are assumed to have the same syntax as 2.0.0.[PAR]",
"The structure assignment for each residue and time is written to an",
"[TT].xpm[tt] matrix file. This file can be visualized with for instance",
"[TT]xv[tt] and can be converted to postscript with [TT]xpm2ps[tt].",
"Individual chains are separated by light grey lines in the [TT].xpm[tt] and",
"postscript files.",
"The number of residues with each secondary structure type and the",
"total secondary structure ([TT]-sss[tt]) count as a function of",
"time are also written to file ([TT]-sc[tt]).[PAR]",
"Solvent accessible surface (SAS) per residue can be calculated, both in",
"absolute values (A^2) and in fractions of the maximal accessible",
"surface of a residue. The maximal accessible surface is defined as",
"the accessible surface of a residue in a chain of glycines.",
"[BB]Note[bb] that the program [gmx-sas] can also compute SAS",
"and that is more efficient.[PAR]",
"Finally, this program can dump the secondary structure in a special file",
"[TT]ssdump.dat[tt] for usage in the program [gmx-chi]. Together",
"these two programs can be used to analyze dihedral properties as a",
"function of secondary structure type."
};
static gmx_bool bVerbose;
static const char *ss_string = "HEBT";
static int dsspVersion = 2;
t_pargs pa[] = {
{ "-v", FALSE, etBOOL, {&bVerbose},
"HIDDENGenerate miles of useless information" },
{ "-sss", FALSE, etSTR, {&ss_string},
"Secondary structures for structure count"},
{ "-ver", FALSE, etINT, {&dsspVersion},
"DSSP major version. Syntax changed with version 2"}
};
t_trxstatus *status;
FILE *tapein;
FILE *ss, *acc, *fTArea, *tmpf;
const char *fnSCount, *fnArea, *fnTArea, *fnAArea;
const char *leg[] = { "Phobic", "Phylic" };
t_topology top;
int ePBC;
t_atoms *atoms;
t_matrix mat;
int nres, nr0, naccr, nres_plus_separators;
gmx_bool *bPhbres, bDoAccSurf;
real t;
int i, j, natoms, nframe = 0;
matrix box = {{0}};
int gnx;
char *grpnm, *ss_str;
atom_id *index;
rvec *xp, *x;
int *average_area;
real **accr, *accr_ptr = NULL, *av_area, *norm_av_area;
char pdbfile[32], tmpfile[32], title[256];
char dssp[256];
const char *dptr;
output_env_t oenv;
gmx_rmpbc_t gpbc = NULL;
t_filenm fnm[] = {
{ efTRX, "-f", NULL, ffREAD },
{ efTPS, NULL, NULL, ffREAD },
{ efNDX, NULL, NULL, ffOPTRD },
{ efDAT, "-ssdump", "ssdump", ffOPTWR },
{ efMAP, "-map", "ss", ffLIBRD },
{ efXPM, "-o", "ss", ffWRITE },
{ efXVG, "-sc", "scount", ffWRITE },
{ efXPM, "-a", "area", ffOPTWR },
{ efXVG, "-ta", "totarea", ffOPTWR },
{ efXVG, "-aa", "averarea", ffOPTWR }
};
#define NFILE asize(fnm)
if (!parse_common_args(&argc, argv,
PCA_CAN_TIME | PCA_CAN_VIEW | PCA_TIME_UNIT,
NFILE, fnm, asize(pa), pa, asize(desc), desc, 0, NULL, &oenv))
{
return 0;
}
fnSCount = opt2fn("-sc", NFILE, fnm);
fnArea = opt2fn_null("-a", NFILE, fnm);
fnTArea = opt2fn_null("-ta", NFILE, fnm);
fnAArea = opt2fn_null("-aa", NFILE, fnm);
bDoAccSurf = (fnArea || fnTArea || fnAArea);
read_tps_conf(ftp2fn(efTPS, NFILE, fnm), title, &top, &ePBC, &xp, NULL, box, FALSE);
atoms = &(top.atoms);
check_oo(atoms);
bPhbres = bPhobics(atoms);
get_index(atoms, ftp2fn_null(efNDX, NFILE, fnm), 1, &gnx, &index, &grpnm);
nres = 0;
nr0 = -1;
for (i = 0; (i < gnx); i++)
{
if (atoms->atom[index[i]].resind != nr0)
{
nr0 = atoms->atom[index[i]].resind;
nres++;
}
}
fprintf(stderr, "There are %d residues in your selected group\n", nres);
strcpy(pdbfile, "ddXXXXXX");
gmx_tmpnam(pdbfile);
if ((tmpf = fopen(pdbfile, "w")) == NULL)
{
sprintf(pdbfile, "%ctmp%cfilterXXXXXX", DIR_SEPARATOR, DIR_SEPARATOR);
gmx_tmpnam(pdbfile);
if ((tmpf = fopen(pdbfile, "w")) == NULL)
{
gmx_fatal(FARGS, "Can not open tmp file %s", pdbfile);
}
}
else
{
fclose(tmpf);
}
strcpy(tmpfile, "ddXXXXXX");
gmx_tmpnam(tmpfile);
if ((tmpf = fopen(tmpfile, "w")) == NULL)
{
sprintf(tmpfile, "%ctmp%cfilterXXXXXX", DIR_SEPARATOR, DIR_SEPARATOR);
gmx_tmpnam(tmpfile);
if ((tmpf = fopen(tmpfile, "w")) == NULL)
{
gmx_fatal(FARGS, "Can not open tmp file %s", tmpfile);
}
}
else
{
fclose(tmpf);
}
if ((dptr = getenv("DSSP")) == NULL)
{
dptr = "/usr/local/bin/dssp";
}
if (!gmx_fexist(dptr))
{
gmx_fatal(FARGS, "DSSP executable (%s) does not exist (use setenv DSSP)",
dptr);
}
if (dsspVersion >= 2)
{
if (dsspVersion > 2)
{
printf("\nWARNING: You use DSSP version %d, which is not explicitly\nsupported by do_dssp. Assuming version 2 syntax.\n\n", dsspVersion);
}
sprintf(dssp, "%s -i %s -o %s > /dev/null %s",
dptr, pdbfile, tmpfile, bVerbose ? "" : "2> /dev/null");
}
else
{
sprintf(dssp, "%s %s %s %s > /dev/null %s",
dptr, bDoAccSurf ? "" : "-na", pdbfile, tmpfile, bVerbose ? "" : "2> /dev/null");
}
fprintf(stderr, "dssp cmd='%s'\n", dssp);
if (fnTArea)
{
fTArea = xvgropen(fnTArea, "Solvent Accessible Surface Area",
output_env_get_xvgr_tlabel(oenv), "Area (nm\\S2\\N)", oenv);
xvgr_legend(fTArea, 2, leg, oenv);
}
else
{
fTArea = NULL;
}
mat.map = NULL;
mat.nmap = readcmap(opt2fn("-map", NFILE, fnm), &(mat.map));
natoms = read_first_x(oenv, &status, ftp2fn(efTRX, NFILE, fnm), &t, &x, box);
if (natoms > atoms->nr)
{
gmx_fatal(FARGS, "\nTrajectory does not match topology!");
}
if (gnx > natoms)
{
gmx_fatal(FARGS, "\nTrajectory does not match selected group!");
}
snew(average_area, atoms->nres);
snew(av_area, atoms->nres);
snew(norm_av_area, atoms->nres);
accr = NULL;
naccr = 0;
gpbc = gmx_rmpbc_init(&top.idef, ePBC, natoms);
do
{
t = output_env_conv_time(oenv, t);
if (bDoAccSurf && nframe >= naccr)
{
naccr += 10;
srenew(accr, naccr);
for (i = naccr-10; i < naccr; i++)
{
snew(accr[i], 2*atoms->nres-1);
}
}
gmx_rmpbc(gpbc, natoms, box, x);
tapein = gmx_ffopen(pdbfile, "w");
write_pdbfile_indexed(tapein, NULL, atoms, x, ePBC, box, ' ', -1, gnx, index, NULL, TRUE);
gmx_ffclose(tapein);
if (0 != system(dssp))
{
gmx_fatal(FARGS, "Failed to execute command: %s\n",
"Try specifying your dssp version with the -ver option.", dssp);
}
/* strip_dssp returns the number of lines found in the dssp file, i.e.
* the number of residues plus the separator lines */
if (bDoAccSurf)
{
accr_ptr = accr[nframe];
}
nres_plus_separators = strip_dssp(tmpfile, nres, bPhbres, t,
accr_ptr, fTArea, &mat, average_area, oenv);
remove(tmpfile);
remove(pdbfile);
nframe++;
}
while (read_next_x(oenv, status, &t, x, box));
fprintf(stderr, "\n");
close_trj(status);
if (fTArea)
{
xvgrclose(fTArea);
}
gmx_rmpbc_done(gpbc);
prune_ss_legend(&mat);
ss = opt2FILE("-o", NFILE, fnm, "w");
mat.flags = 0;
write_xpm_m(ss, mat);
gmx_ffclose(ss);
if (opt2bSet("-ssdump", NFILE, fnm))
{
ss = opt2FILE("-ssdump", NFILE, fnm, "w");
snew(ss_str, nres+1);
fprintf(ss, "%d\n", nres);
for (j = 0; j < mat.nx; j++)
{
for (i = 0; (i < mat.ny); i++)
{
ss_str[i] = mat.map[mat.matrix[j][i]].code.c1;
}
ss_str[i] = '\0';
fprintf(ss, "%s\n", ss_str);
}
gmx_ffclose(ss);
sfree(ss_str);
}
analyse_ss(fnSCount, &mat, ss_string, oenv);
if (bDoAccSurf)
{
write_sas_mat(fnArea, accr, nframe, nres_plus_separators, &mat);
for (i = 0; i < atoms->nres; i++)
{
av_area[i] = (average_area[i] / (real)nframe);
}
norm_acc(atoms, nres, av_area, norm_av_area);
if (fnAArea)
{
acc = xvgropen(fnAArea, "Average Accessible Area",
"Residue", "A\\S2", oenv);
for (i = 0; (i < nres); i++)
{
fprintf(acc, "%5d %10g %10g\n", i+1, av_area[i], norm_av_area[i]);
}
xvgrclose(acc);
}
}
view_all(oenv, NFILE, fnm);
return 0;
}
int gmx_rms(int argc, char *argv[])
{
const char *desc[] =
{
"[THISMODULE] compares two structures by computing the root mean square",
"deviation (RMSD), the size-independent [GRK]rho[grk] similarity parameter",
"([TT]rho[tt]) or the scaled [GRK]rho[grk] ([TT]rhosc[tt]), ",
"see Maiorov & Crippen, Proteins [BB]22[bb], 273 (1995).",
"This is selected by [TT]-what[tt].[PAR]"
"Each structure from a trajectory ([TT]-f[tt]) is compared to a",
"reference structure. The reference structure",
"is taken from the structure file ([TT]-s[tt]).[PAR]",
"With option [TT]-mir[tt] also a comparison with the mirror image of",
"the reference structure is calculated.",
"This is useful as a reference for 'significant' values, see",
"Maiorov & Crippen, Proteins [BB]22[bb], 273 (1995).[PAR]",
"Option [TT]-prev[tt] produces the comparison with a previous frame",
"the specified number of frames ago.[PAR]",
"Option [TT]-m[tt] produces a matrix in [TT].xpm[tt] format of",
"comparison values of each structure in the trajectory with respect to",
"each other structure. This file can be visualized with for instance",
"[TT]xv[tt] and can be converted to postscript with [gmx-xpm2ps].[PAR]",
"Option [TT]-fit[tt] controls the least-squares fitting of",
"the structures on top of each other: complete fit (rotation and",
"translation), translation only, or no fitting at all.[PAR]",
"Option [TT]-mw[tt] controls whether mass weighting is done or not.",
"If you select the option (default) and ",
"supply a valid [TT].tpr[tt] file masses will be taken from there, ",
"otherwise the masses will be deduced from the [TT]atommass.dat[tt] file in",
"[TT]GMXLIB[tt]. This is fine for proteins, but not",
"necessarily for other molecules. A default mass of 12.011 amu (carbon)",
"is assigned to unknown atoms. You can check whether this happend by",
"turning on the [TT]-debug[tt] flag and inspecting the log file.[PAR]",
"With [TT]-f2[tt], the 'other structures' are taken from a second",
"trajectory, this generates a comparison matrix of one trajectory",
"versus the other.[PAR]",
"Option [TT]-bin[tt] does a binary dump of the comparison matrix.[PAR]",
"Option [TT]-bm[tt] produces a matrix of average bond angle deviations",
"analogously to the [TT]-m[tt] option. Only bonds between atoms in the",
"comparison group are considered."
};
static gmx_bool bPBC = TRUE, bFitAll = TRUE, bSplit = FALSE;
static gmx_bool bDeltaLog = FALSE;
static int prev = 0, freq = 1, freq2 = 1, nlevels = 80, avl = 0;
static real rmsd_user_max = -1, rmsd_user_min = -1, bond_user_max = -1,
bond_user_min = -1, delta_maxy = 0.0;
/* strings and things for selecting difference method */
enum
{
ewSel, ewRMSD, ewRho, ewRhoSc, ewNR
};
int ewhat;
const char *what[ewNR + 1] =
{ NULL, "rmsd", "rho", "rhosc", NULL };
const char *whatname[ewNR] =
{ NULL, "RMSD", "Rho", "Rho sc" };
const char *whatlabel[ewNR] =
{ NULL, "RMSD (nm)", "Rho", "Rho sc" };
const char *whatxvgname[ewNR] =
{ NULL, "RMSD", "\\8r\\4", "\\8r\\4\\ssc\\N" };
const char *whatxvglabel[ewNR] =
{ NULL, "RMSD (nm)", "\\8r\\4", "\\8r\\4\\ssc\\N" };
/* strings and things for fitting methods */
enum
{
efSel, efFit, efReset, efNone, efNR
};
int efit;
const char *fit[efNR + 1] =
{ NULL, "rot+trans", "translation", "none", NULL };
const char *fitgraphlabel[efNR + 1] =
{ NULL, "lsq fit", "translational fit", "no fit" };
static int nrms = 1;
static gmx_bool bMassWeighted = TRUE;
t_pargs pa[] =
{
{ "-what", FALSE, etENUM,
{ what }, "Structural difference measure" },
{ "-pbc", FALSE, etBOOL,
{ &bPBC }, "PBC check" },
{ "-fit", FALSE, etENUM,
{ fit }, "Fit to reference structure" },
{ "-prev", FALSE, etINT,
{ &prev }, "Compare with previous frame" },
{ "-split", FALSE, etBOOL,
{ &bSplit }, "Split graph where time is zero" },
{ "-fitall", FALSE, etBOOL,
{ &bFitAll }, "HIDDENFit all pairs of structures in matrix" },
{ "-skip", FALSE, etINT,
{ &freq }, "Only write every nr-th frame to matrix" },
{ "-skip2", FALSE, etINT,
{ &freq2 }, "Only write every nr-th frame to matrix" },
{ "-max", FALSE, etREAL,
{ &rmsd_user_max }, "Maximum level in comparison matrix" },
{ "-min", FALSE, etREAL,
{ &rmsd_user_min }, "Minimum level in comparison matrix" },
{ "-bmax", FALSE, etREAL,
{ &bond_user_max }, "Maximum level in bond angle matrix" },
{ "-bmin", FALSE, etREAL,
{ &bond_user_min }, "Minimum level in bond angle matrix" },
{ "-mw", FALSE, etBOOL,
{ &bMassWeighted }, "Use mass weighting for superposition" },
{ "-nlevels", FALSE, etINT,
{ &nlevels }, "Number of levels in the matrices" },
{ "-ng", FALSE, etINT,
{ &nrms }, "Number of groups to compute RMS between" },
{ "-dlog", FALSE, etBOOL,
{ &bDeltaLog },
"HIDDENUse a log x-axis in the delta t matrix" },
{ "-dmax", FALSE, etREAL,
{ &delta_maxy }, "HIDDENMaximum level in delta matrix" },
{ "-aver", FALSE, etINT,
{ &avl },
"HIDDENAverage over this distance in the RMSD matrix" }
};
int natoms_trx, natoms_trx2, natoms;
int i, j, k, m, teller, teller2, tel_mat, tel_mat2;
#define NFRAME 5000
int maxframe = NFRAME, maxframe2 = NFRAME;
real t, *w_rls, *w_rms, *w_rls_m = NULL, *w_rms_m = NULL;
gmx_bool bNorm, bAv, bFreq2, bFile2, bMat, bBond, bDelta, bMirror, bMass;
gmx_bool bFit, bReset;
t_topology top;
int ePBC;
t_iatom *iatom = NULL;
matrix box;
rvec *x, *xp, *xm = NULL, **mat_x = NULL, **mat_x2, *mat_x2_j = NULL, vec1,
vec2;
t_trxstatus *status;
char buf[256], buf2[256];
int ncons = 0;
FILE *fp;
real rlstot = 0, **rls, **rlsm = NULL, *time, *time2, *rlsnorm = NULL,
**rmsd_mat = NULL, **bond_mat = NULL, *axis, *axis2, *del_xaxis,
*del_yaxis, rmsd_max, rmsd_min, rmsd_avg, bond_max, bond_min, ang;
real **rmsdav_mat = NULL, av_tot, weight, weight_tot;
real **delta = NULL, delta_max, delta_scalex = 0, delta_scaley = 0,
*delta_tot;
int delta_xsize = 0, del_lev = 100, mx, my, abs_my;
gmx_bool bA1, bA2, bPrev, bTop, *bInMat = NULL;
int ifit, *irms, ibond = 0, *ind_bond1 = NULL, *ind_bond2 = NULL, n_ind_m =
0;
atom_id *ind_fit, **ind_rms, *ind_m = NULL, *rev_ind_m = NULL, *ind_rms_m =
NULL;
char *gn_fit, **gn_rms;
t_rgb rlo, rhi;
output_env_t oenv;
gmx_rmpbc_t gpbc = NULL;
t_filenm fnm[] =
{
{ efTPS, NULL, NULL, ffREAD },
{ efTRX, "-f", NULL, ffREAD },
{ efTRX, "-f2", NULL, ffOPTRD },
{ efNDX, NULL, NULL, ffOPTRD },
{ efXVG, NULL, "rmsd", ffWRITE },
{ efXVG, "-mir", "rmsdmir", ffOPTWR },
{ efXVG, "-a", "avgrp", ffOPTWR },
{ efXVG, "-dist", "rmsd-dist", ffOPTWR },
{ efXPM, "-m", "rmsd", ffOPTWR },
{ efDAT, "-bin", "rmsd", ffOPTWR },
{ efXPM, "-bm", "bond", ffOPTWR }
};
#define NFILE asize(fnm)
if (!parse_common_args(&argc, argv, PCA_CAN_TIME | PCA_TIME_UNIT | PCA_CAN_VIEW
| PCA_BE_NICE, NFILE, fnm, asize(pa), pa, asize(desc), desc, 0, NULL,
&oenv))
{
return 0;
}
/* parse enumerated options: */
ewhat = nenum(what);
if (ewhat == ewRho || ewhat == ewRhoSc)
{
please_cite(stdout, "Maiorov95");
}
efit = nenum(fit);
bFit = efit == efFit;
bReset = efit == efReset;
if (bFit)
{
bReset = TRUE; /* for fit, reset *must* be set */
}
else
{
bFitAll = FALSE;
}
/* mark active cmdline options */
bMirror = opt2bSet("-mir", NFILE, fnm); /* calc RMSD vs mirror of ref. */
bFile2 = opt2bSet("-f2", NFILE, fnm);
bMat = opt2bSet("-m", NFILE, fnm);
bBond = opt2bSet("-bm", NFILE, fnm);
bDelta = (delta_maxy > 0); /* calculate rmsd vs delta t matrix from *
* your RMSD matrix (hidden option */
bNorm = opt2bSet("-a", NFILE, fnm);
bFreq2 = opt2parg_bSet("-skip2", asize(pa), pa);
if (freq <= 0)
{
fprintf(stderr, "The number of frames to skip is <= 0. "
"Writing out all frames.\n\n");
freq = 1;
}
if (!bFreq2)
{
freq2 = freq;
}
else if (bFile2 && freq2 <= 0)
{
fprintf(stderr,
"The number of frames to skip in second trajectory is <= 0.\n"
" Writing out all frames.\n\n");
freq2 = 1;
}
bPrev = (prev > 0);
if (bPrev)
{
prev = abs(prev);
if (freq != 1)
{
fprintf(stderr, "WARNING: option -skip also applies to -prev\n");
}
}
if (bFile2 && !bMat && !bBond)
{
fprintf(
stderr,
"WARNING: second trajectory (-f2) useless when not calculating matrix (-m/-bm),\n"
" will not read from %s\n", opt2fn("-f2", NFILE,
fnm));
bFile2 = FALSE;
}
if (bDelta)
{
bMat = TRUE;
if (bFile2)
{
fprintf(stderr,
"WARNING: second trajectory (-f2) useless when making delta matrix,\n"
" will not read from %s\n", opt2fn("-f2",
NFILE, fnm));
bFile2 = FALSE;
}
}
bTop = read_tps_conf(ftp2fn(efTPS, NFILE, fnm), buf, &top, &ePBC, &xp,
NULL, box, TRUE);
snew(w_rls, top.atoms.nr);
snew(w_rms, top.atoms.nr);
if (!bTop && bBond)
{
fprintf(stderr,
"WARNING: Need a run input file for bond angle matrix,\n"
" will not calculate bond angle matrix.\n");
bBond = FALSE;
}
if (bReset)
{
fprintf(stderr, "Select group for %s fit\n", bFit ? "least squares"
: "translational");
get_index(&(top.atoms), ftp2fn_null(efNDX, NFILE, fnm), 1, &ifit,
&ind_fit, &gn_fit);
}
else
{
ifit = 0;
}
if (bReset)
{
if (bFit && ifit < 3)
{
gmx_fatal(FARGS, "Need >= 3 points to fit!\n" );
}
bMass = FALSE;
for (i = 0; i < ifit; i++)
{
if (bMassWeighted)
{
w_rls[ind_fit[i]] = top.atoms.atom[ind_fit[i]].m;
}
else
{
w_rls[ind_fit[i]] = 1;
}
bMass = bMass || (top.atoms.atom[ind_fit[i]].m != 0);
}
if (!bMass)
{
fprintf(stderr, "All masses in the fit group are 0, using masses of 1\n");
for (i = 0; i < ifit; i++)
{
w_rls[ind_fit[i]] = 1;
}
}
}
if (bMat || bBond)
{
nrms = 1;
}
snew(gn_rms, nrms);
snew(ind_rms, nrms);
snew(irms, nrms);
fprintf(stderr, "Select group%s for %s calculation\n",
(nrms > 1) ? "s" : "", whatname[ewhat]);
get_index(&(top.atoms), ftp2fn_null(efNDX, NFILE, fnm),
nrms, irms, ind_rms, gn_rms);
if (bNorm)
{
snew(rlsnorm, irms[0]);
}
snew(rls, nrms);
for (j = 0; j < nrms; j++)
{
snew(rls[j], maxframe);
}
if (bMirror)
{
snew(rlsm, nrms);
for (j = 0; j < nrms; j++)
{
snew(rlsm[j], maxframe);
}
}
snew(time, maxframe);
for (j = 0; j < nrms; j++)
{
bMass = FALSE;
for (i = 0; i < irms[j]; i++)
{
if (bMassWeighted)
{
w_rms[ind_rms[j][i]] = top.atoms.atom[ind_rms[j][i]].m;
}
else
{
w_rms[ind_rms[j][i]] = 1.0;
}
bMass = bMass || (top.atoms.atom[ind_rms[j][i]].m != 0);
}
if (!bMass)
{
fprintf(stderr, "All masses in group %d are 0, using masses of 1\n", j);
for (i = 0; i < irms[j]; i++)
{
w_rms[ind_rms[j][i]] = 1;
}
}
}
/* Prepare reference frame */
if (bPBC)
{
gpbc = gmx_rmpbc_init(&top.idef, ePBC, top.atoms.nr);
gmx_rmpbc(gpbc, top.atoms.nr, box, xp);
}
if (bReset)
{
reset_x(ifit, ind_fit, top.atoms.nr, NULL, xp, w_rls);
}
if (bMirror)
{
/* generate reference structure mirror image: */
snew(xm, top.atoms.nr);
for (i = 0; i < top.atoms.nr; i++)
{
copy_rvec(xp[i], xm[i]);
xm[i][XX] = -xm[i][XX];
}
}
if (ewhat == ewRhoSc)
{
norm_princ(&top.atoms, ifit, ind_fit, top.atoms.nr, xp);
}
/* read first frame */
natoms_trx = read_first_x(oenv, &status, opt2fn("-f", NFILE, fnm), &t, &x, box);
if (natoms_trx != top.atoms.nr)
{
fprintf(stderr,
"\nWARNING: topology has %d atoms, whereas trajectory has %d\n",
top.atoms.nr, natoms_trx);
}
natoms = min(top.atoms.nr, natoms_trx);
if (bMat || bBond || bPrev)
{
snew(mat_x, NFRAME);
if (bPrev)
{
/* With -prev we use all atoms for simplicity */
n_ind_m = natoms;
}
else
{
/* Check which atoms we need (fit/rms) */
snew(bInMat, natoms);
for (i = 0; i < ifit; i++)
{
bInMat[ind_fit[i]] = TRUE;
}
n_ind_m = ifit;
for (i = 0; i < irms[0]; i++)
{
if (!bInMat[ind_rms[0][i]])
{
bInMat[ind_rms[0][i]] = TRUE;
n_ind_m++;
}
}
}
/* Make an index of needed atoms */
snew(ind_m, n_ind_m);
snew(rev_ind_m, natoms);
j = 0;
for (i = 0; i < natoms; i++)
{
if (bPrev || bInMat[i])
{
ind_m[j] = i;
rev_ind_m[i] = j;
j++;
}
}
snew(w_rls_m, n_ind_m);
snew(ind_rms_m, irms[0]);
snew(w_rms_m, n_ind_m);
for (i = 0; i < ifit; i++)
{
w_rls_m[rev_ind_m[ind_fit[i]]] = w_rls[ind_fit[i]];
}
for (i = 0; i < irms[0]; i++)
{
ind_rms_m[i] = rev_ind_m[ind_rms[0][i]];
w_rms_m[ind_rms_m[i]] = w_rms[ind_rms[0][i]];
}
sfree(bInMat);
}
if (bBond)
{
ncons = 0;
for (k = 0; k < F_NRE; k++)
{
if (IS_CHEMBOND(k))
{
iatom = top.idef.il[k].iatoms;
ncons += top.idef.il[k].nr/3;
}
}
fprintf(stderr, "Found %d bonds in topology\n", ncons);
snew(ind_bond1, ncons);
snew(ind_bond2, ncons);
ibond = 0;
for (k = 0; k < F_NRE; k++)
{
if (IS_CHEMBOND(k))
{
iatom = top.idef.il[k].iatoms;
ncons = top.idef.il[k].nr/3;
for (i = 0; i < ncons; i++)
{
bA1 = FALSE;
bA2 = FALSE;
for (j = 0; j < irms[0]; j++)
{
if (iatom[3*i+1] == ind_rms[0][j])
{
bA1 = TRUE;
}
if (iatom[3*i+2] == ind_rms[0][j])
{
bA2 = TRUE;
}
}
if (bA1 && bA2)
{
ind_bond1[ibond] = rev_ind_m[iatom[3*i+1]];
ind_bond2[ibond] = rev_ind_m[iatom[3*i+2]];
ibond++;
}
}
}
}
fprintf(stderr, "Using %d bonds for bond angle matrix\n", ibond);
if (ibond == 0)
{
gmx_fatal(FARGS, "0 bonds found");
}
}
/* start looping over frames: */
tel_mat = 0;
teller = 0;
do
{
if (bPBC)
{
gmx_rmpbc(gpbc, natoms, box, x);
}
if (bReset)
{
reset_x(ifit, ind_fit, natoms, NULL, x, w_rls);
}
if (ewhat == ewRhoSc)
{
norm_princ(&top.atoms, ifit, ind_fit, natoms, x);
}
if (bFit)
{
/*do the least squares fit to original structure*/
do_fit(natoms, w_rls, xp, x);
}
if (teller % freq == 0)
{
/* keep frame for matrix calculation */
if (bMat || bBond || bPrev)
{
if (tel_mat >= NFRAME)
{
srenew(mat_x, tel_mat+1);
}
snew(mat_x[tel_mat], n_ind_m);
for (i = 0; i < n_ind_m; i++)
{
copy_rvec(x[ind_m[i]], mat_x[tel_mat][i]);
}
}
tel_mat++;
}
/*calculate energy of root_least_squares*/
if (bPrev)
{
j = tel_mat-prev-1;
if (j < 0)
{
j = 0;
}
for (i = 0; i < n_ind_m; i++)
{
copy_rvec(mat_x[j][i], xp[ind_m[i]]);
}
if (bReset)
{
reset_x(ifit, ind_fit, natoms, NULL, xp, w_rls);
}
if (bFit)
{
do_fit(natoms, w_rls, x, xp);
}
}
for (j = 0; (j < nrms); j++)
{
rls[j][teller] =
calc_similar_ind(ewhat != ewRMSD, irms[j], ind_rms[j], w_rms, x, xp);
}
if (bNorm)
{
for (j = 0; (j < irms[0]); j++)
{
rlsnorm[j] +=
calc_similar_ind(ewhat != ewRMSD, 1, &(ind_rms[0][j]), w_rms, x, xp);
}
}
if (bMirror)
{
if (bFit)
{
/*do the least squares fit to mirror of original structure*/
do_fit(natoms, w_rls, xm, x);
}
for (j = 0; j < nrms; j++)
{
rlsm[j][teller] =
calc_similar_ind(ewhat != ewRMSD, irms[j], ind_rms[j], w_rms, x, xm);
}
}
time[teller] = output_env_conv_time(oenv, t);
teller++;
if (teller >= maxframe)
{
maxframe += NFRAME;
srenew(time, maxframe);
for (j = 0; (j < nrms); j++)
{
srenew(rls[j], maxframe);
}
if (bMirror)
{
for (j = 0; (j < nrms); j++)
{
srenew(rlsm[j], maxframe);
}
}
}
}
while (read_next_x(oenv, status, &t, x, box));
close_trj(status);
if (bFile2)
{
snew(time2, maxframe2);
fprintf(stderr, "\nWill read second trajectory file\n");
snew(mat_x2, NFRAME);
natoms_trx2 =
read_first_x(oenv, &status, opt2fn("-f2", NFILE, fnm), &t, &x, box);
if (natoms_trx2 != natoms_trx)
{
gmx_fatal(FARGS,
"Second trajectory (%d atoms) does not match the first one"
" (%d atoms)", natoms_trx2, natoms_trx);
}
tel_mat2 = 0;
teller2 = 0;
do
{
if (bPBC)
{
gmx_rmpbc(gpbc, natoms, box, x);
}
if (bReset)
{
reset_x(ifit, ind_fit, natoms, NULL, x, w_rls);
}
if (ewhat == ewRhoSc)
{
norm_princ(&top.atoms, ifit, ind_fit, natoms, x);
}
if (bFit)
{
/*do the least squares fit to original structure*/
do_fit(natoms, w_rls, xp, x);
}
if (teller2 % freq2 == 0)
{
/* keep frame for matrix calculation */
if (bMat)
{
if (tel_mat2 >= NFRAME)
{
srenew(mat_x2, tel_mat2+1);
}
snew(mat_x2[tel_mat2], n_ind_m);
for (i = 0; i < n_ind_m; i++)
{
copy_rvec(x[ind_m[i]], mat_x2[tel_mat2][i]);
}
}
tel_mat2++;
}
time2[teller2] = output_env_conv_time(oenv, t);
teller2++;
if (teller2 >= maxframe2)
{
maxframe2 += NFRAME;
srenew(time2, maxframe2);
}
}
while (read_next_x(oenv, status, &t, x, box));
close_trj(status);
}
else
{
mat_x2 = mat_x;
time2 = time;
tel_mat2 = tel_mat;
freq2 = freq;
}
gmx_rmpbc_done(gpbc);
if (bMat || bBond)
{
/* calculate RMS matrix */
fprintf(stderr, "\n");
if (bMat)
{
fprintf(stderr, "Building %s matrix, %dx%d elements\n",
whatname[ewhat], tel_mat, tel_mat2);
snew(rmsd_mat, tel_mat);
}
if (bBond)
{
fprintf(stderr, "Building bond angle matrix, %dx%d elements\n",
tel_mat, tel_mat2);
snew(bond_mat, tel_mat);
}
snew(axis, tel_mat);
snew(axis2, tel_mat2);
rmsd_max = 0;
if (bFile2)
{
rmsd_min = 1e10;
}
else
{
rmsd_min = 0;
}
rmsd_avg = 0;
bond_max = 0;
bond_min = 1e10;
for (j = 0; j < tel_mat2; j++)
{
axis2[j] = time2[freq2*j];
}
if (bDelta)
{
if (bDeltaLog)
{
delta_scalex = 8.0/log(2.0);
delta_xsize = (int)(log(tel_mat/2)*delta_scalex+0.5)+1;
}
else
{
delta_xsize = tel_mat/2;
}
delta_scaley = 1.0/delta_maxy;
snew(delta, delta_xsize);
for (j = 0; j < delta_xsize; j++)
{
snew(delta[j], del_lev+1);
}
if (avl > 0)
{
snew(rmsdav_mat, tel_mat);
for (j = 0; j < tel_mat; j++)
{
snew(rmsdav_mat[j], tel_mat);
}
}
}
if (bFitAll)
{
snew(mat_x2_j, natoms);
}
for (i = 0; i < tel_mat; i++)
{
axis[i] = time[freq*i];
fprintf(stderr, "\r element %5d; time %5.2f ", i, axis[i]);
if (bMat)
{
snew(rmsd_mat[i], tel_mat2);
}
if (bBond)
{
snew(bond_mat[i], tel_mat2);
}
for (j = 0; j < tel_mat2; j++)
{
if (bFitAll)
{
for (k = 0; k < n_ind_m; k++)
{
copy_rvec(mat_x2[j][k], mat_x2_j[k]);
}
do_fit(n_ind_m, w_rls_m, mat_x[i], mat_x2_j);
}
else
{
mat_x2_j = mat_x2[j];
}
if (bMat)
{
if (bFile2 || (i < j))
{
rmsd_mat[i][j] =
calc_similar_ind(ewhat != ewRMSD, irms[0], ind_rms_m,
w_rms_m, mat_x[i], mat_x2_j);
if (rmsd_mat[i][j] > rmsd_max)
{
rmsd_max = rmsd_mat[i][j];
}
if (rmsd_mat[i][j] < rmsd_min)
{
rmsd_min = rmsd_mat[i][j];
}
rmsd_avg += rmsd_mat[i][j];
}
else
{
rmsd_mat[i][j] = rmsd_mat[j][i];
}
}
if (bBond)
{
if (bFile2 || (i <= j))
{
ang = 0.0;
for (m = 0; m < ibond; m++)
{
rvec_sub(mat_x[i][ind_bond1[m]], mat_x[i][ind_bond2[m]], vec1);
rvec_sub(mat_x2_j[ind_bond1[m]], mat_x2_j[ind_bond2[m]], vec2);
ang += acos(cos_angle(vec1, vec2));
}
bond_mat[i][j] = ang*180.0/(M_PI*ibond);
if (bond_mat[i][j] > bond_max)
{
bond_max = bond_mat[i][j];
}
if (bond_mat[i][j] < bond_min)
{
bond_min = bond_mat[i][j];
}
}
else
{
bond_mat[i][j] = bond_mat[j][i];
}
}
}
}
if (bFile2)
{
rmsd_avg /= tel_mat*tel_mat2;
}
else
{
rmsd_avg /= tel_mat*(tel_mat - 1)/2;
}
if (bMat && (avl > 0))
{
rmsd_max = 0.0;
rmsd_min = 0.0;
rmsd_avg = 0.0;
for (j = 0; j < tel_mat-1; j++)
{
for (i = j+1; i < tel_mat; i++)
{
av_tot = 0;
weight_tot = 0;
for (my = -avl; my <= avl; my++)
{
if ((j+my >= 0) && (j+my < tel_mat))
{
abs_my = abs(my);
for (mx = -avl; mx <= avl; mx++)
{
if ((i+mx >= 0) && (i+mx < tel_mat))
{
weight = (real)(avl+1-max(abs(mx), abs_my));
av_tot += weight*rmsd_mat[i+mx][j+my];
weight_tot += weight;
}
}
}
}
rmsdav_mat[i][j] = av_tot/weight_tot;
rmsdav_mat[j][i] = rmsdav_mat[i][j];
if (rmsdav_mat[i][j] > rmsd_max)
{
rmsd_max = rmsdav_mat[i][j];
}
}
}
rmsd_mat = rmsdav_mat;
}
if (bMat)
{
fprintf(stderr, "\n%s: Min %f, Max %f, Avg %f\n",
whatname[ewhat], rmsd_min, rmsd_max, rmsd_avg);
rlo.r = 1; rlo.g = 1; rlo.b = 1;
rhi.r = 0; rhi.g = 0; rhi.b = 0;
if (rmsd_user_max != -1)
{
rmsd_max = rmsd_user_max;
}
if (rmsd_user_min != -1)
{
rmsd_min = rmsd_user_min;
}
if ((rmsd_user_max != -1) || (rmsd_user_min != -1))
{
fprintf(stderr, "Min and Max value set to resp. %f and %f\n",
rmsd_min, rmsd_max);
}
sprintf(buf, "%s %s matrix", gn_rms[0], whatname[ewhat]);
write_xpm(opt2FILE("-m", NFILE, fnm, "w"), 0, buf, whatlabel[ewhat],
output_env_get_time_label(oenv), output_env_get_time_label(oenv), tel_mat, tel_mat2,
axis, axis2, rmsd_mat, rmsd_min, rmsd_max, rlo, rhi, &nlevels);
/* Print the distribution of RMSD values */
if (opt2bSet("-dist", NFILE, fnm))
{
low_rmsd_dist(opt2fn("-dist", NFILE, fnm), rmsd_max, tel_mat, rmsd_mat, oenv);
}
if (bDelta)
{
snew(delta_tot, delta_xsize);
for (j = 0; j < tel_mat-1; j++)
{
for (i = j+1; i < tel_mat; i++)
{
mx = i-j;
if (mx < tel_mat/2)
{
if (bDeltaLog)
{
mx = (int)(log(mx)*delta_scalex+0.5);
}
my = (int)(rmsd_mat[i][j]*delta_scaley*del_lev+0.5);
delta_tot[mx] += 1.0;
if ((rmsd_mat[i][j] >= 0) && (rmsd_mat[i][j] <= delta_maxy))
{
delta[mx][my] += 1.0;
}
}
}
}
delta_max = 0;
for (i = 0; i < delta_xsize; i++)
{
if (delta_tot[i] > 0.0)
{
delta_tot[i] = 1.0/delta_tot[i];
for (j = 0; j <= del_lev; j++)
{
delta[i][j] *= delta_tot[i];
if (delta[i][j] > delta_max)
{
delta_max = delta[i][j];
}
}
}
}
fprintf(stderr, "Maximum in delta matrix: %f\n", delta_max);
snew(del_xaxis, delta_xsize);
snew(del_yaxis, del_lev+1);
for (i = 0; i < delta_xsize; i++)
{
del_xaxis[i] = axis[i]-axis[0];
}
for (i = 0; i < del_lev+1; i++)
{
del_yaxis[i] = delta_maxy*i/del_lev;
}
sprintf(buf, "%s %s vs. delta t", gn_rms[0], whatname[ewhat]);
fp = gmx_ffopen("delta.xpm", "w");
write_xpm(fp, 0, buf, "density", output_env_get_time_label(oenv), whatlabel[ewhat],
delta_xsize, del_lev+1, del_xaxis, del_yaxis,
delta, 0.0, delta_max, rlo, rhi, &nlevels);
gmx_ffclose(fp);
}
if (opt2bSet("-bin", NFILE, fnm))
{
/* NB: File must be binary if we use fwrite */
fp = ftp2FILE(efDAT, NFILE, fnm, "wb");
for (i = 0; i < tel_mat; i++)
{
if (fwrite(rmsd_mat[i], sizeof(**rmsd_mat), tel_mat2, fp) != tel_mat2)
{
gmx_fatal(FARGS, "Error writing to output file");
}
}
gmx_ffclose(fp);
}
}
if (bBond)
{
fprintf(stderr, "\nMin. angle: %f, Max. angle: %f\n", bond_min, bond_max);
if (bond_user_max != -1)
{
bond_max = bond_user_max;
}
if (bond_user_min != -1)
{
bond_min = bond_user_min;
}
if ((bond_user_max != -1) || (bond_user_min != -1))
{
fprintf(stderr, "Bond angle Min and Max set to:\n"
"Min. angle: %f, Max. angle: %f\n", bond_min, bond_max);
}
rlo.r = 1; rlo.g = 1; rlo.b = 1;
rhi.r = 0; rhi.g = 0; rhi.b = 0;
sprintf(buf, "%s av. bond angle deviation", gn_rms[0]);
write_xpm(opt2FILE("-bm", NFILE, fnm, "w"), 0, buf, "degrees",
output_env_get_time_label(oenv), output_env_get_time_label(oenv), tel_mat, tel_mat2,
axis, axis2, bond_mat, bond_min, bond_max, rlo, rhi, &nlevels);
}
}
bAv = opt2bSet("-a", NFILE, fnm);
/* Write the RMSD's to file */
if (!bPrev)
{
sprintf(buf, "%s", whatxvgname[ewhat]);
}
else
{
sprintf(buf, "%s with frame %g %s ago", whatxvgname[ewhat],
time[prev*freq]-time[0], output_env_get_time_label(oenv));
}
fp = xvgropen(opt2fn("-o", NFILE, fnm), buf, output_env_get_xvgr_tlabel(oenv),
whatxvglabel[ewhat], oenv);
if (output_env_get_print_xvgr_codes(oenv))
{
fprintf(fp, "@ subtitle \"%s%s after %s%s%s\"\n",
(nrms == 1) ? "" : "of ", gn_rms[0], fitgraphlabel[efit],
bFit ? " to " : "", bFit ? gn_fit : "");
}
if (nrms != 1)
{
xvgr_legend(fp, nrms, (const char**)gn_rms, oenv);
}
for (i = 0; (i < teller); i++)
{
if (bSplit && i > 0 &&
abs(time[bPrev ? freq*i : i]/output_env_get_time_factor(oenv)) < 1e-5)
{
fprintf(fp, "&\n");
}
fprintf(fp, "%12.7f", time[bPrev ? freq*i : i]);
for (j = 0; (j < nrms); j++)
{
fprintf(fp, " %12.7f", rls[j][i]);
if (bAv)
{
rlstot += rls[j][i];
}
}
fprintf(fp, "\n");
}
gmx_ffclose(fp);
if (bMirror)
{
/* Write the mirror RMSD's to file */
sprintf(buf, "%s with Mirror", whatxvgname[ewhat]);
sprintf(buf2, "Mirror %s", whatxvglabel[ewhat]);
fp = xvgropen(opt2fn("-mir", NFILE, fnm), buf, output_env_get_xvgr_tlabel(oenv),
buf2, oenv);
if (nrms == 1)
{
if (output_env_get_print_xvgr_codes(oenv))
{
fprintf(fp, "@ subtitle \"of %s after lsq fit to mirror of %s\"\n",
gn_rms[0], gn_fit);
}
}
else
{
if (output_env_get_print_xvgr_codes(oenv))
{
fprintf(fp, "@ subtitle \"after lsq fit to mirror %s\"\n", gn_fit);
}
xvgr_legend(fp, nrms, (const char**)gn_rms, oenv);
}
for (i = 0; (i < teller); i++)
{
if (bSplit && i > 0 && abs(time[i]) < 1e-5)
{
fprintf(fp, "&\n");
}
fprintf(fp, "%12.7f", time[i]);
for (j = 0; (j < nrms); j++)
{
fprintf(fp, " %12.7f", rlsm[j][i]);
}
fprintf(fp, "\n");
}
gmx_ffclose(fp);
}
if (bAv)
{
sprintf(buf, "Average %s", whatxvgname[ewhat]);
sprintf(buf2, "Average %s", whatxvglabel[ewhat]);
fp = xvgropen(opt2fn("-a", NFILE, fnm), buf, "Residue", buf2, oenv);
for (j = 0; (j < nrms); j++)
{
fprintf(fp, "%10d %10g\n", j, rlstot/teller);
}
gmx_ffclose(fp);
}
if (bNorm)
{
fp = xvgropen("aver.xvg", gn_rms[0], "Residue", whatxvglabel[ewhat], oenv);
for (j = 0; (j < irms[0]); j++)
{
fprintf(fp, "%10d %10g\n", j, rlsnorm[j]/teller);
}
gmx_ffclose(fp);
}
do_view(oenv, opt2fn_null("-a", NFILE, fnm), "-graphtype bar");
do_view(oenv, opt2fn("-o", NFILE, fnm), NULL);
do_view(oenv, opt2fn_null("-mir", NFILE, fnm), NULL);
do_view(oenv, opt2fn_null("-m", NFILE, fnm), NULL);
do_view(oenv, opt2fn_null("-bm", NFILE, fnm), NULL);
do_view(oenv, opt2fn_null("-dist", NFILE, fnm), NULL);
return 0;
}
int main(int argc,char *argv[])
{
static char *desc[] = {
"do_dssp ",
"reads a trajectory file and computes the secondary structure for",
"each time frame ",
"calling the dssp program. If you do not have the dssp program,",
"get it. do_dssp assumes that the dssp executable is",
"/usr/local/bin/dssp. If this is not the case, then you should",
"set an environment variable [BB]DSSP[bb] pointing to the dssp",
"executable, e.g.: [PAR]",
"[TT]setenv DSSP /opt/dssp/bin/dssp[tt][PAR]",
"The structure assignment for each residue and time is written to an",
"[TT].xpm[tt] matrix file. This file can be visualized with for instance",
"[TT]xv[tt] and can be converted to postscript with [TT]xpm2ps[tt].",
"The number of residues with each secondary structure type and the",
"total secondary structure ([TT]-sss[tt]) count as a function of",
"time are also written to file ([TT]-sc[tt]).[PAR]",
"Solvent accessible surface (SAS) per residue can be calculated, both in",
"absolute values (A^2) and in fractions of the maximal accessible",
"surface of a residue. The maximal accessible surface is defined as",
"the accessible surface of a residue in a chain of glycines.",
"[BB]Note[bb] that the program [TT]g_sas[tt] can also compute SAS",
"and that is more efficient.[PAR]",
"Finally, this program can dump the secondary structure in a special file",
"[TT]ssdump.dat[tt] for usage in the program [TT]g_chi[tt]. Together",
"these two programs can be used to analyze dihedral properties as a",
"function of secondary structure type."
};
static bool bVerbose;
static char *ss_string="HEBT";
t_pargs pa[] = {
{ "-v", FALSE, etBOOL, {&bVerbose},
"HIDDENGenerate miles of useless information" },
{ "-sss", FALSE, etSTR, {&ss_string},
"Secondary structures for structure count"}
};
int status;
FILE *tapein;
FILE *ss,*acc,*fTArea,*tmpf;
char *fnSCount,*fnArea,*fnTArea,*fnAArea;
char *leg[] = { "Phobic", "Phylic" };
t_topology top;
int ePBC;
t_atoms *atoms;
t_matrix mat;
int nres,nr0,naccr;
bool *bPhbres,bDoAccSurf;
real t;
int i,j,natoms,nframe=0;
matrix box;
int gnx;
char *grpnm,*ss_str;
atom_id *index;
rvec *xp,*x;
int *average_area;
real **accr,*av_area, *norm_av_area;
char pdbfile[32],tmpfile[32],title[256];
char dssp[256],*dptr;
t_filenm fnm[] = {
{ efTRX, "-f", NULL, ffREAD },
{ efTPS, NULL, NULL, ffREAD },
{ efNDX, NULL, NULL, ffOPTRD },
{ efDAT, "-ssdump", "ssdump", ffOPTWR },
{ efMAP, "-map", "ss", ffLIBRD },
{ efXPM, "-o", "ss", ffWRITE },
{ efXVG, "-sc", "scount", ffWRITE },
{ efXPM, "-a", "area", ffOPTWR },
{ efXVG, "-ta", "totarea", ffOPTWR },
{ efXVG, "-aa", "averarea",ffOPTWR }
};
#define NFILE asize(fnm)
CopyRight(stderr,argv[0]);
parse_common_args(&argc,argv,PCA_CAN_TIME | PCA_CAN_VIEW | PCA_TIME_UNIT | PCA_BE_NICE ,
NFILE,fnm, asize(pa),pa, asize(desc),desc,0,NULL);
fnSCount= opt2fn("-sc",NFILE,fnm);
fnArea = opt2fn_null("-a", NFILE,fnm);
fnTArea = opt2fn_null("-ta",NFILE,fnm);
fnAArea = opt2fn_null("-aa",NFILE,fnm);
bDoAccSurf=(fnArea || fnTArea || fnAArea);
read_tps_conf(ftp2fn(efTPS,NFILE,fnm),title,&top,&ePBC,&xp,NULL,box,FALSE);
atoms=&(top.atoms);
check_oo(atoms);
bPhbres=bPhobics(atoms);
get_index(atoms,ftp2fn_null(efNDX,NFILE,fnm),1,&gnx,&index,&grpnm);
nres=0;
nr0=-1;
for(i=0; (i<gnx); i++) {
if (atoms->atom[index[i]].resnr != nr0) {
nr0=atoms->atom[index[i]].resnr;
nres++;
}
}
fprintf(stderr,"There are %d residues in your selected group\n",nres);
strcpy(pdbfile,"ddXXXXXX");
gmx_tmpnam(pdbfile);
if ((tmpf = fopen(pdbfile,"w")) == NULL) {
sprintf(pdbfile,"%ctmp%cfilterXXXXXX",DIR_SEPARATOR,DIR_SEPARATOR);
gmx_tmpnam(pdbfile);
if ((tmpf = fopen(pdbfile,"w")) == NULL)
gmx_fatal(FARGS,"Can not open tmp file %s",pdbfile);
}
else
fclose(tmpf);
strcpy(tmpfile,"ddXXXXXX");
gmx_tmpnam(tmpfile);
if ((tmpf = fopen(tmpfile,"w")) == NULL) {
sprintf(tmpfile,"%ctmp%cfilterXXXXXX",DIR_SEPARATOR,DIR_SEPARATOR);
gmx_tmpnam(tmpfile);
if ((tmpf = fopen(tmpfile,"w")) == NULL)
gmx_fatal(FARGS,"Can not open tmp file %s",tmpfile);
}
else
fclose(tmpf);
if ((dptr=getenv("DSSP")) == NULL)
dptr="/usr/local/bin/dssp";
if (!fexist(dptr))
gmx_fatal(FARGS,"DSSP executable (%s) does not exist (use setenv DSSP)",
dptr);
sprintf(dssp,"%s %s %s %s > /dev/null %s",
dptr,bDoAccSurf?"":"-na",pdbfile,tmpfile,bVerbose?"":"2> /dev/null");
if (bVerbose)
fprintf(stderr,"dssp cmd='%s'\n",dssp);
if (fnTArea) {
fTArea=xvgropen(fnTArea,"Solvent Accessible Surface Area",
xvgr_tlabel(),"Area (nm\\S2\\N)");
xvgr_legend(fTArea,2,leg);
} else
fTArea=NULL;
mat.map=NULL;
mat.nmap=getcmap(libopen(opt2fn("-map",NFILE,fnm)),
opt2fn("-map",NFILE,fnm),&(mat.map));
natoms=read_first_x(&status,ftp2fn(efTRX,NFILE,fnm),&t,&x,box);
if (natoms > atoms->nr)
gmx_fatal(FARGS,"\nTrajectory does not match topology!");
if (gnx > natoms)
gmx_fatal(FARGS,"\nTrajectory does not match selected group!");
snew(average_area,atoms->nres+10);
snew(av_area,atoms->nres+10);
snew(norm_av_area,atoms->nres+10);
accr=NULL;
naccr=0;
do {
t = convert_time(t);
if (nframe>=naccr) {
naccr+=10;
srenew(accr,naccr);
for(i=naccr-10; i<naccr; i++)
snew(accr[i],atoms->nres+10);
}
rm_pbc(&(top.idef),ePBC,natoms,box,x,x);
tapein=ffopen(pdbfile,"w");
write_pdbfile_indexed(tapein,NULL,atoms,x,ePBC,box,0,-1,gnx,index);
fclose(tapein);
#ifdef GMX_NO_SYSTEM
printf("Warning-- No calls to system(3) supported on this platform.");
printf("Warning-- Skipping execution of 'system(\"%s\")'.", dssp);
exit(1);
#else
if(0 != system(dssp))
{
gmx_fatal(FARGS,"Failed to execute command: %s",dssp);
}
#endif
strip_dssp(tmpfile,nres,bPhbres,t,
accr[nframe],fTArea,&mat,average_area);
remove(tmpfile);
remove(pdbfile);
nframe++;
} while(read_next_x(status,&t,natoms,x,box));
fprintf(stderr,"\n");
close_trj(status);
if (fTArea)
ffclose(fTArea);
prune_ss_legend(&mat);
ss=opt2FILE("-o",NFILE,fnm,"w");
write_xpm_m(ss,mat);
ffclose(ss);
if (opt2bSet("-ssdump",NFILE,fnm)) {
snew(ss_str,nres+1);
for(i=0; (i<nres); i++)
ss_str[i] = mat.map[mat.matrix[0][i]].code.c1;
ss_str[i] = '\0';
ss = opt2FILE("-ssdump",NFILE,fnm,"w");
fprintf(ss,"%d\n%s\n",nres,ss_str);
fclose(ss);
sfree(ss_str);
}
analyse_ss(fnSCount,&mat,ss_string);
if (bDoAccSurf) {
write_sas_mat(fnArea,accr,nframe,nres,&mat);
for(i=0; i<atoms->nres; i++)
av_area[i] = (average_area[i] / (real)nframe);
norm_acc(atoms, nres, av_area, norm_av_area);
if (fnAArea) {
acc=xvgropen(fnAArea,"Average Accessible Area",
"Residue","A\\S2");
for(i=0; (i<nres); i++)
fprintf(acc,"%5d %10g %10g\n",i+1,av_area[i], norm_av_area[i]);
ffclose(acc);
}
}
view_all(NFILE, fnm);
thanx(stderr);
return 0;
}
int gmx_genpr(int argc,char *argv[])
{
const char *desc[] = {
"[TT]genrestr[tt] produces an include file for a topology containing",
"a list of atom numbers and three force constants for the",
"[IT]x[it]-, [IT]y[it]-, and [IT]z[it]-direction. A single isotropic force constant may",
"be given on the command line instead of three components.[PAR]",
"WARNING: position restraints only work for the one molecule at a time.",
"Position restraints are interactions within molecules, therefore",
"they should be included within the correct [TT][ moleculetype ][tt]",
"block in the topology. Since the atom numbers in every moleculetype",
"in the topology start at 1 and the numbers in the input file for",
"[TT]genrestr[tt] number consecutively from 1, [TT]genrestr[tt] will only",
"produce a useful file for the first molecule.[PAR]",
"The [TT]-of[tt] option produces an index file that can be used for",
"freezing atoms. In this case, the input file must be a [TT].pdb[tt] file.[PAR]",
"With the [TT]-disre[tt] option, half a matrix of distance restraints",
"is generated instead of position restraints. With this matrix, that",
"one typically would apply to C[GRK]alpha[grk] atoms in a protein, one can",
"maintain the overall conformation of a protein without tieing it to",
"a specific position (as with position restraints)."
};
static rvec fc={1000.0,1000.0,1000.0};
static real freeze_level = 0.0;
static real disre_dist = 0.1;
static real disre_frac = 0.0;
static real disre_up2 = 1.0;
static gmx_bool bDisre=FALSE;
static gmx_bool bConstr=FALSE;
static real cutoff = -1.0;
t_pargs pa[] = {
{ "-fc", FALSE, etRVEC, {fc},
"force constants (kJ/mol nm^2)" },
{ "-freeze", FALSE, etREAL, {&freeze_level},
"if the [TT]-of[tt] option or this one is given an index file will be written containing atom numbers of all atoms that have a B-factor less than the level given here" },
{ "-disre", FALSE, etBOOL, {&bDisre},
"Generate a distance restraint matrix for all the atoms in index" },
{ "-disre_dist", FALSE, etREAL, {&disre_dist},
"Distance range around the actual distance for generating distance restraints" },
{ "-disre_frac", FALSE, etREAL, {&disre_frac},
"Fraction of distance to be used as interval rather than a fixed distance. If the fraction of the distance that you specify here is less than the distance given in the previous option, that one is used instead." },
{ "-disre_up2", FALSE, etREAL, {&disre_up2},
"Distance between upper bound for distance restraints, and the distance at which the force becomes constant (see manual)" },
{ "-cutoff", FALSE, etREAL, {&cutoff},
"Only generate distance restraints for atoms pairs within cutoff (nm)" },
{ "-constr", FALSE, etBOOL, {&bConstr},
"Generate a constraint matrix rather than distance restraints. Constraints of type 2 will be generated that do generate exclusions." }
};
#define npargs asize(pa)
output_env_t oenv;
t_atoms *atoms=NULL;
int i,j,k;
FILE *out;
int igrp;
real d,dd,lo,hi;
atom_id *ind_grp;
const char *xfn,*nfn;
char *gn_grp;
char title[STRLEN];
matrix box;
gmx_bool bFreeze;
rvec dx,*x=NULL,*v=NULL;
t_filenm fnm[] = {
{ efSTX, "-f", NULL, ffREAD },
{ efNDX, "-n", NULL, ffOPTRD },
{ efITP, "-o", "posre", ffWRITE },
{ efNDX, "-of", "freeze", ffOPTWR }
};
#define NFILE asize(fnm)
CopyRight(stderr,argv[0]);
parse_common_args(&argc,argv,0,NFILE,fnm,npargs,pa,
asize(desc),desc,0,NULL,&oenv);
bFreeze = opt2bSet("-of",NFILE,fnm) || opt2parg_bSet("-freeze",asize(pa),pa);
bDisre = bDisre || opt2parg_bSet("-disre_dist",npargs,pa);
xfn = opt2fn_null("-f",NFILE,fnm);
nfn = opt2fn_null("-n",NFILE,fnm);
if (( nfn == NULL ) && ( xfn == NULL))
gmx_fatal(FARGS,"no index file and no structure file suplied");
if ((disre_frac < 0) || (disre_frac >= 1))
gmx_fatal(FARGS,"disre_frac should be between 0 and 1");
if (disre_dist < 0)
gmx_fatal(FARGS,"disre_dist should be >= 0");
if (xfn != NULL) {
snew(atoms,1);
get_stx_coordnum(xfn,&(atoms->nr));
init_t_atoms(atoms,atoms->nr,TRUE);
snew(x,atoms->nr);
snew(v,atoms->nr);
fprintf(stderr,"\nReading structure file\n");
read_stx_conf(xfn,title,atoms,x,v,NULL,box);
}
if (bFreeze) {
if (atoms && atoms->pdbinfo)
gmx_fatal(FARGS,"No B-factors in input file %s, use a pdb file next time.",
xfn);
out=opt2FILE("-of",NFILE,fnm,"w");
fprintf(out,"[ freeze ]\n");
for(i=0; (i<atoms->nr); i++) {
if (atoms->pdbinfo[i].bfac <= freeze_level)
fprintf(out,"%d\n",i+1);
}
ffclose(out);
}
else if ((bDisre || bConstr) && x) {
printf("Select group to generate %s matrix from\n",
bConstr ? "constraint" : "distance restraint");
get_index(atoms,nfn,1,&igrp,&ind_grp,&gn_grp);
out=ftp2FILE(efITP,NFILE,fnm,"w");
if (bConstr) {
fprintf(out,"; constraints for %s of %s\n\n",gn_grp,title);
fprintf(out,"[ constraints ]\n");
fprintf(out,";%4s %5s %1s %10s\n","i","j","tp","dist");
}
else {
fprintf(out,"; distance restraints for %s of %s\n\n",gn_grp,title);
fprintf(out,"[ distance_restraints ]\n");
fprintf(out,";%4s %5s %1s %5s %10s %10s %10s %10s %10s\n","i","j","?",
"label","funct","lo","up1","up2","weight");
}
for(i=k=0; i<igrp; i++)
for(j=i+1; j<igrp; j++,k++) {
rvec_sub(x[ind_grp[i]],x[ind_grp[j]],dx);
d = norm(dx);
if (bConstr)
fprintf(out,"%5d %5d %1d %10g\n",ind_grp[i]+1,ind_grp[j]+1,2,d);
else {
if (cutoff < 0 || d < cutoff)
{
if (disre_frac > 0)
dd = min(disre_dist,disre_frac*d);
else
dd = disre_dist;
lo = max(0,d-dd);
hi = d+dd;
fprintf(out,"%5d %5d %1d %5d %10d %10g %10g %10g %10g\n",
ind_grp[i]+1,ind_grp[j]+1,1,k,1,
lo,hi,hi+1,1.0);
}
}
}
ffclose(out);
}
else {
printf("Select group to position restrain\n");
get_index(atoms,nfn,1,&igrp,&ind_grp,&gn_grp);
out=ftp2FILE(efITP,NFILE,fnm,"w");
fprintf(out,"; position restraints for %s of %s\n\n",gn_grp,title);
fprintf(out,"[ position_restraints ]\n");
fprintf(out,";%3s %5s %9s %10s %10s\n","i","funct","fcx","fcy","fcz");
for(i=0; i<igrp; i++)
fprintf(out,"%4d %4d %10g %10g %10g\n",
ind_grp[i]+1,1,fc[XX],fc[YY],fc[ZZ]);
ffclose(out);
}
if (xfn) {
sfree(x);
sfree(v);
}
thanx(stderr);
return 0;
}
int gmx_rmsdist(int argc, char *argv[])
{
const char *desc[] = {
"[THISMODULE] computes the root mean square deviation of atom distances,",
"which has the advantage that no fit is needed like in standard RMS",
"deviation as computed by [gmx-rms].",
"The reference structure is taken from the structure file.",
"The RMSD at time t is calculated as the RMS",
"of the differences in distance between atom-pairs in the reference",
"structure and the structure at time t.[PAR]",
"[THISMODULE] can also produce matrices of the rms distances, rms distances",
"scaled with the mean distance and the mean distances and matrices with",
"NMR averaged distances (1/r^3 and 1/r^6 averaging). Finally, lists",
"of atom pairs with 1/r^3 and 1/r^6 averaged distance below the",
"maximum distance ([TT]-max[tt], which will default to 0.6 in this case)",
"can be generated, by default averaging over equivalent hydrogens",
"(all triplets of hydrogens named \\*[123]). Additionally a list of",
"equivalent atoms can be supplied ([TT]-equiv[tt]), each line containing",
"a set of equivalent atoms specified as residue number and name and",
"atom name; e.g.:[PAR]",
"[TT]HB* 3 SER HB1 3 SER HB2[tt][PAR]",
"Residue and atom names must exactly match those in the structure",
"file, including case. Specifying non-sequential atoms is undefined."
};
int i, teller;
real t;
t_topology top;
int ePBC;
t_atoms *atoms;
matrix box;
rvec *x;
FILE *fp;
t_trxstatus *status;
int isize, gnr = 0;
atom_id *index, *noe_index;
char *grpname;
real **d_r, **d, **dtot, **dtot2, **mean, **rms, **rmsc, *resnr;
real **dtot1_3 = NULL, **dtot1_6 = NULL;
real rmsnow, meanmax, rmsmax, rmscmax;
real max1_3, max1_6;
t_noe_gr *noe_gr = NULL;
t_noe **noe = NULL;
t_rgb rlo, rhi;
gmx_bool bRMS, bScale, bMean, bNOE, bNMR3, bNMR6, bNMR;
static int nlevels = 40;
static real scalemax = -1.0;
static gmx_bool bSumH = TRUE;
static gmx_bool bPBC = TRUE;
output_env_t oenv;
t_pargs pa[] = {
{ "-nlevels", FALSE, etINT, {&nlevels},
"Discretize RMS in this number of levels" },
{ "-max", FALSE, etREAL, {&scalemax},
"Maximum level in matrices" },
{ "-sumh", FALSE, etBOOL, {&bSumH},
"Average distance over equivalent hydrogens" },
{ "-pbc", FALSE, etBOOL, {&bPBC},
"Use periodic boundary conditions when computing distances" }
};
t_filenm fnm[] = {
{ efTRX, "-f", NULL, ffREAD },
{ efTPS, NULL, NULL, ffREAD },
{ efNDX, NULL, NULL, ffOPTRD },
{ efDAT, "-equiv", "equiv", ffOPTRD },
{ efXVG, NULL, "distrmsd", ffWRITE },
{ efXPM, "-rms", "rmsdist", ffOPTWR },
{ efXPM, "-scl", "rmsscale", ffOPTWR },
{ efXPM, "-mean", "rmsmean", ffOPTWR },
{ efXPM, "-nmr3", "nmr3", ffOPTWR },
{ efXPM, "-nmr6", "nmr6", ffOPTWR },
{ efDAT, "-noe", "noe", ffOPTWR },
};
#define NFILE asize(fnm)
if (!parse_common_args(&argc, argv, PCA_CAN_VIEW | PCA_CAN_TIME,
NFILE, fnm, asize(pa), pa, asize(desc), desc, 0, NULL, &oenv))
{
return 0;
}
bRMS = opt2bSet("-rms", NFILE, fnm);
bScale = opt2bSet("-scl", NFILE, fnm);
bMean = opt2bSet("-mean", NFILE, fnm);
bNOE = opt2bSet("-noe", NFILE, fnm);
bNMR3 = opt2bSet("-nmr3", NFILE, fnm);
bNMR6 = opt2bSet("-nmr6", NFILE, fnm);
bNMR = bNMR3 || bNMR6 || bNOE;
max1_3 = 0;
max1_6 = 0;
/* check input */
if (bNOE && scalemax < 0)
{
scalemax = 0.6;
fprintf(stderr, "WARNING: using -noe without -max "
"makes no sense, setting -max to %g\n\n", scalemax);
}
/* get topology and index */
read_tps_conf(ftp2fn(efTPS, NFILE, fnm), &top, &ePBC, &x, NULL, box, FALSE);
if (!bPBC)
{
ePBC = epbcNONE;
}
atoms = &(top.atoms);
get_index(atoms, ftp2fn_null(efNDX, NFILE, fnm), 1, &isize, &index, &grpname);
/* initialize arrays */
snew(d, isize);
snew(dtot, isize);
snew(dtot2, isize);
if (bNMR)
{
snew(dtot1_3, isize);
snew(dtot1_6, isize);
}
snew(mean, isize);
snew(rms, isize);
snew(rmsc, isize);
snew(d_r, isize);
snew(resnr, isize);
for (i = 0; (i < isize); i++)
{
snew(d[i], isize);
snew(dtot[i], isize);
snew(dtot2[i], isize);
if (bNMR)
{
snew(dtot1_3[i], isize);
snew(dtot1_6[i], isize);
}
snew(mean[i], isize);
snew(rms[i], isize);
snew(rmsc[i], isize);
snew(d_r[i], isize);
resnr[i] = i+1;
}
/*set box type*/
calc_dist(isize, index, x, ePBC, box, d_r);
sfree(x);
/*open output files*/
fp = xvgropen(ftp2fn(efXVG, NFILE, fnm), "RMS Deviation", "Time (ps)", "RMSD (nm)",
oenv);
if (output_env_get_print_xvgr_codes(oenv))
{
fprintf(fp, "@ subtitle \"of distances between %s atoms\"\n", grpname);
}
/*do a first step*/
read_first_x(oenv, &status, ftp2fn(efTRX, NFILE, fnm), &t, &x, box);
teller = 0;
do
{
calc_dist_tot(isize, index, x, ePBC, box, d, dtot, dtot2, bNMR, dtot1_3, dtot1_6);
rmsnow = rms_diff(isize, d, d_r);
fprintf(fp, "%g %g\n", t, rmsnow);
teller++;
}
while (read_next_x(oenv, status, &t, x, box));
fprintf(stderr, "\n");
xvgrclose(fp);
close_trj(status);
calc_rms(isize, teller, dtot, dtot2, rms, &rmsmax, rmsc, &rmscmax, mean, &meanmax);
fprintf(stderr, "rmsmax = %g, rmscmax = %g\n", rmsmax, rmscmax);
if (bNMR)
{
calc_nmr(isize, teller, dtot1_3, dtot1_6, &max1_3, &max1_6);
}
if (scalemax > -1.0)
{
rmsmax = scalemax;
rmscmax = scalemax;
meanmax = scalemax;
max1_3 = scalemax;
max1_6 = scalemax;
}
if (bNOE)
{
/* make list of noe atom groups */
snew(noe_index, isize+1);
snew(noe_gr, isize);
gnr = analyze_noe_equivalent(opt2fn_null("-equiv", NFILE, fnm),
atoms, isize, index, bSumH, noe_index, noe_gr);
fprintf(stdout, "Found %d non-equivalent atom-groups in %d atoms\n",
gnr, isize);
/* make half matrix of of noe-group distances from atom distances */
snew(noe, gnr);
for (i = 0; i < gnr; i++)
{
snew(noe[i], gnr);
}
calc_noe(isize, noe_index, dtot1_3, dtot1_6, gnr, noe);
}
rlo.r = 1.0, rlo.g = 1.0, rlo.b = 1.0;
rhi.r = 0.0, rhi.g = 0.0, rhi.b = 0.0;
if (bRMS)
{
write_xpm(opt2FILE("-rms", NFILE, fnm, "w"), 0,
"RMS of distance", "RMS (nm)", "Atom Index", "Atom Index",
isize, isize, resnr, resnr, rms, 0.0, rmsmax, rlo, rhi, &nlevels);
}
if (bScale)
{
write_xpm(opt2FILE("-scl", NFILE, fnm, "w"), 0,
"Relative RMS", "RMS", "Atom Index", "Atom Index",
isize, isize, resnr, resnr, rmsc, 0.0, rmscmax, rlo, rhi, &nlevels);
}
if (bMean)
{
write_xpm(opt2FILE("-mean", NFILE, fnm, "w"), 0,
"Mean Distance", "Distance (nm)", "Atom Index", "Atom Index",
isize, isize, resnr, resnr, mean, 0.0, meanmax, rlo, rhi, &nlevels);
}
if (bNMR3)
{
write_xpm(opt2FILE("-nmr3", NFILE, fnm, "w"), 0, "1/r^3 averaged distances",
"Distance (nm)", "Atom Index", "Atom Index",
isize, isize, resnr, resnr, dtot1_3, 0.0, max1_3, rlo, rhi, &nlevels);
}
if (bNMR6)
{
write_xpm(opt2FILE("-nmr6", NFILE, fnm, "w"), 0, "1/r^6 averaged distances",
"Distance (nm)", "Atom Index", "Atom Index",
isize, isize, resnr, resnr, dtot1_6, 0.0, max1_6, rlo, rhi, &nlevels);
}
if (bNOE)
{
write_noe(opt2FILE("-noe", NFILE, fnm, "w"), gnr, noe, noe_gr, scalemax);
}
do_view(oenv, ftp2fn(efXVG, NFILE, fnm), NULL);
return 0;
}
int gmx_cluster(int argc,char *argv[])
{
static char *desc[] = {
"g_cluster can cluster structures with several different methods.",
"Distances between structures can be determined from a trajectory",
"or read from an XPM matrix file with the [TT]-dm[tt] option.",
"RMS deviation after fitting or RMS deviation of atom-pair distances",
"can be used to define the distance between structures.[PAR]",
"single linkage: add a structure to a cluster when its distance to any",
"element of the cluster is less than [TT]cutoff[tt].[PAR]",
"Jarvis Patrick: add a structure to a cluster when this structure",
"and a structure in the cluster have each other as neighbors and",
"they have a least [TT]P[tt] neighbors in common. The neighbors",
"of a structure are the M closest structures or all structures within",
"[TT]cutoff[tt].[PAR]",
"Monte Carlo: reorder the RMSD matrix using Monte Carlo.[PAR]",
"diagonalization: diagonalize the RMSD matrix.[PAR]"
"gromos: use algorithm as described in Daura [IT]et al.[it]",
"([IT]Angew. Chem. Int. Ed.[it] [BB]1999[bb], [IT]38[it], pp 236-240).",
"Count number of neighbors using cut-off, take structure with",
"largest number of neighbors with all its neighbors as cluster",
"and eleminate it from the pool of clusters. Repeat for remaining",
"structures in pool.[PAR]",
"When the clustering algorithm assigns each structure to exactly one",
"cluster (single linkage, Jarvis Patrick and gromos) and a trajectory",
"file is supplied, the structure with",
"the smallest average distance to the others or the average structure",
"or all structures for each cluster will be written to a trajectory",
"file. When writing all structures, separate numbered files are made",
"for each cluster.[PAR]"
"Two output files are always written:[BR]",
"[TT]-o[tt] writes the RMSD values in the upper left half of the matrix",
"and a graphical depiction of the clusters in the lower right half",
"When [TT]-minstruct[tt] = 1 the graphical depiction is black",
"when two structures are in the same cluster.",
"When [TT]-minstruct[tt] > 1 different colors will be used for each",
"cluster.[BR]",
"[TT]-g[tt] writes information on the options used and a detailed list",
"of all clusters and their members.[PAR]",
"Additionally, a number of optional output files can be written:[BR]",
"[TT]-dist[tt] writes the RMSD distribution.[BR]",
"[TT]-ev[tt] writes the eigenvectors of the RMSD matrix",
"diagonalization.[BR]",
"[TT]-sz[tt] writes the cluster sizes.[BR]",
"[TT]-tr[tt] writes a matrix of the number transitions between",
"cluster pairs.[BR]",
"[TT]-ntr[tt] writes the total number of transitions to or from",
"each cluster.[BR]",
"[TT]-clid[tt] writes the cluster number as a function of time.[BR]",
"[TT]-cl[tt] writes average (with option [TT]-av[tt]) or central",
"structure of each cluster or writes numbered files with cluster members",
"for a selected set of clusters (with option [TT]-wcl[tt], depends on",
"[TT]-nst[tt] and [TT]-rmsmin[tt]).[BR]",
};
FILE *fp,*log;
int i,i1,i2,j,nf,nrms;
matrix box;
rvec *xtps,*usextps,*x1,**xx=NULL;
char *fn,*trx_out_fn;
t_clusters clust;
t_mat *rms;
real *eigval;
t_topology top;
int ePBC;
t_atoms useatoms;
t_matrix *readmat;
real *tmp;
int isize=0,ifsize=0,iosize=0;
atom_id *index=NULL, *fitidx, *outidx;
char *grpname;
real rmsd,**d1,**d2,*time,time_invfac,*mass=NULL;
char buf[STRLEN],buf1[80],title[STRLEN];
bool bAnalyze,bUseRmsdCut,bJP_RMSD=FALSE,bReadMat,bReadTraj;
int method,ncluster=0;
static char *methodname[] = {
NULL, "linkage", "jarvis-patrick","monte-carlo",
"diagonalization", "gromos", NULL
};
enum { m_null, m_linkage, m_jarvis_patrick,
m_monte_carlo, m_diagonalize, m_gromos, m_nr };
/* Set colors for plotting: white = zero RMS, black = maximum */
static t_rgb rlo_top = { 1.0, 1.0, 1.0 };
static t_rgb rhi_top = { 0.0, 0.0, 0.0 };
static t_rgb rlo_bot = { 1.0, 1.0, 1.0 };
static t_rgb rhi_bot = { 0.0, 0.0, 1.0 };
static int nlevels=40,skip=1;
static real scalemax=-1.0,rmsdcut=0.1,rmsmin=0.0;
static bool bRMSdist=FALSE,bBinary=FALSE,bAverage=FALSE,bFit=TRUE;
static int niter=10000,seed=1993,write_ncl=0,write_nst=1,minstruct=1;
static real kT=1e-3;
static int M=10,P=3;
t_pargs pa[] = {
{ "-dista", FALSE, etBOOL, {&bRMSdist},
"Use RMSD of distances instead of RMS deviation" },
{ "-nlevels",FALSE,etINT, {&nlevels},
"Discretize RMSD matrix in # levels" },
{ "-cutoff",FALSE, etREAL, {&rmsdcut},
"RMSD cut-off (nm) for two structures to be neighbor" },
{ "-fit", FALSE, etBOOL, {&bFit},
"Use least squares fitting before RMSD calculation" },
{ "-max", FALSE, etREAL, {&scalemax},
"Maximum level in RMSD matrix" },
{ "-skip", FALSE, etINT, {&skip},
"Only analyze every nr-th frame" },
{ "-av", FALSE, etBOOL, {&bAverage},
"Write average iso middle structure for each cluster" },
{ "-wcl", FALSE, etINT, {&write_ncl},
"Write all structures for first # clusters to numbered files" },
{ "-nst", FALSE, etINT, {&write_nst},
"Only write all structures if more than # per cluster" },
{ "-rmsmin",FALSE, etREAL, {&rmsmin},
"minimum rms difference with rest of cluster for writing structures" },
{ "-method",FALSE, etENUM, {methodname},
"Method for cluster determination" },
{ "-minstruct", FALSE, etINT, {&minstruct},
"Minimum number of structures in cluster for coloring in the xpm file" },
{ "-binary",FALSE, etBOOL, {&bBinary},
"Treat the RMSD matrix as consisting of 0 and 1, where the cut-off "
"is given by -cutoff" },
{ "-M", FALSE, etINT, {&M},
"Number of nearest neighbors considered for Jarvis-Patrick algorithm, "
"0 is use cutoff" },
{ "-P", FALSE, etINT, {&P},
"Number of identical nearest neighbors required to form a cluster" },
{ "-seed", FALSE, etINT, {&seed},
"Random number seed for Monte Carlo clustering algorithm" },
{ "-niter", FALSE, etINT, {&niter},
"Number of iterations for MC" },
{ "-kT", FALSE, etREAL, {&kT},
"Boltzmann weighting factor for Monte Carlo optimization "
"(zero turns off uphill steps)" }
};
t_filenm fnm[] = {
{ efTRX, "-f", NULL, ffOPTRD },
{ efTPS, "-s", NULL, ffOPTRD },
{ efNDX, NULL, NULL, ffOPTRD },
{ efXPM, "-dm", "rmsd", ffOPTRD },
{ efXPM, "-o", "rmsd-clust", ffWRITE },
{ efLOG, "-g", "cluster", ffWRITE },
{ efXVG, "-dist", "rmsd-dist", ffOPTWR },
{ efXVG, "-ev", "rmsd-eig", ffOPTWR },
{ efXVG, "-sz", "clust-size", ffOPTWR},
{ efXPM, "-tr", "clust-trans",ffOPTWR},
{ efXVG, "-ntr", "clust-trans",ffOPTWR},
{ efXVG, "-clid", "clust-id.xvg",ffOPTWR},
{ efTRX, "-cl", "clusters.pdb", ffOPTWR }
};
#define NFILE asize(fnm)
CopyRight(stderr,argv[0]);
parse_common_args(&argc,argv,PCA_CAN_VIEW | PCA_CAN_TIME | PCA_TIME_UNIT | PCA_BE_NICE,
NFILE,fnm,asize(pa),pa,asize(desc),desc,0,NULL);
/* parse options */
bReadMat = opt2bSet("-dm",NFILE,fnm);
bReadTraj = opt2bSet("-f",NFILE,fnm) || !bReadMat;
if ( opt2parg_bSet("-av",asize(pa),pa) ||
opt2parg_bSet("-wcl",asize(pa),pa) ||
opt2parg_bSet("-nst",asize(pa),pa) ||
opt2parg_bSet("-rmsmin",asize(pa),pa) ||
opt2bSet("-cl",NFILE,fnm) )
trx_out_fn = opt2fn("-cl",NFILE,fnm);
else
trx_out_fn = NULL;
if (bReadMat && time_factor()!=1) {
fprintf(stderr,
"\nWarning: assuming the time unit in %s is %s\n",
opt2fn("-dm",NFILE,fnm),time_unit());
}
if (trx_out_fn && !bReadTraj)
fprintf(stderr,"\nWarning: "
"cannot write cluster structures without reading trajectory\n"
" ignoring option -cl %s\n", trx_out_fn);
method=1;
while ( method < m_nr && strcasecmp(methodname[0], methodname[method])!=0 )
method++;
if (method == m_nr)
gmx_fatal(FARGS,"Invalid method");
bAnalyze = (method == m_linkage || method == m_jarvis_patrick ||
method == m_gromos );
/* Open log file */
log = ftp2FILE(efLOG,NFILE,fnm,"w");
fprintf(stderr,"Using %s method for clustering\n",methodname[0]);
fprintf(log,"Using %s method for clustering\n",methodname[0]);
/* check input and write parameters to log file */
bUseRmsdCut = FALSE;
if (method == m_jarvis_patrick) {
bJP_RMSD = (M == 0) || opt2parg_bSet("-cutoff",asize(pa),pa);
if ((M<0) || (M == 1))
gmx_fatal(FARGS,"M (%d) must be 0 or larger than 1",M);
if (M < 2) {
sprintf(buf1,"Will use P=%d and RMSD cutoff (%g)",P,rmsdcut);
bUseRmsdCut = TRUE;
} else {
if (P >= M)
gmx_fatal(FARGS,"Number of neighbors required (P) must be less than M");
if (bJP_RMSD) {
sprintf(buf1,"Will use P=%d, M=%d and RMSD cutoff (%g)",P,M,rmsdcut);
bUseRmsdCut = TRUE;
} else
sprintf(buf1,"Will use P=%d, M=%d",P,M);
}
ffprintf1(stderr,log,buf,"%s for determining the neighbors\n\n",buf1);
} else /* method != m_jarvis */
bUseRmsdCut = ( bBinary || method == m_linkage || method == m_gromos );
if (bUseRmsdCut && method != m_jarvis_patrick)
fprintf(log,"Using RMSD cutoff %g nm\n",rmsdcut);
if ( method==m_monte_carlo )
fprintf(log,"Using %d iterations\n",niter);
if (skip < 1)
gmx_fatal(FARGS,"skip (%d) should be >= 1",skip);
/* get input */
if (bReadTraj) {
/* don't read mass-database as masses (and top) are not used */
read_tps_conf(ftp2fn(efTPS,NFILE,fnm),buf,&top,&ePBC,&xtps,NULL,box,
bAnalyze);
fprintf(stderr,"\nSelect group for least squares fit%s:\n",
bReadMat?"":" and RMSD calculation");
get_index(&(top.atoms),ftp2fn_null(efNDX,NFILE,fnm),
1,&ifsize,&fitidx,&grpname);
if (trx_out_fn) {
fprintf(stderr,"\nSelect group for output:\n");
get_index(&(top.atoms),ftp2fn_null(efNDX,NFILE,fnm),
1,&iosize,&outidx,&grpname);
/* merge and convert both index groups: */
/* first copy outidx to index. let outidx refer to elements in index */
snew(index,iosize);
isize = iosize;
for(i=0; i<iosize; i++) {
index[i]=outidx[i];
outidx[i]=i;
}
/* now lookup elements from fitidx in index, add them if necessary
and also let fitidx refer to elements in index */
for(i=0; i<ifsize; i++) {
j=0;
while (j<isize && index[j]!=fitidx[i])
j++;
if (j>=isize) {
/* slow this way, but doesn't matter much */
isize++;
srenew(index,isize);
}
index[j]=fitidx[i];
fitidx[i]=j;
}
} else { /* !trx_out_fn */
isize = ifsize;
snew(index, isize);
for(i=0; i<ifsize; i++) {
index[i]=fitidx[i];
fitidx[i]=i;
}
}
}
/* Initiate arrays */
snew(d1,isize);
snew(d2,isize);
for(i=0; (i<isize); i++) {
snew(d1[i],isize);
snew(d2[i],isize);
}
if (bReadTraj) {
/* Loop over first coordinate file */
fn = opt2fn("-f",NFILE,fnm);
xx = read_whole_trj(fn,isize,index,skip,&nf,&time);
convert_times(nf, time);
if (!bRMSdist || bAnalyze) {
/* Center all frames on zero */
snew(mass,isize);
for(i=0; i<ifsize; i++)
mass[fitidx[i]] = top.atoms.atom[index[fitidx[i]]].m;
if (bFit)
for(i=0; i<nf; i++)
reset_x(ifsize,fitidx,isize,NULL,xx[i],mass);
}
}
if (bReadMat) {
fprintf(stderr,"Reading rms distance matrix ");
read_xpm_matrix(opt2fn("-dm",NFILE,fnm),&readmat);
fprintf(stderr,"\n");
if (readmat[0].nx != readmat[0].ny)
gmx_fatal(FARGS,"Matrix (%dx%d) is not square",
readmat[0].nx,readmat[0].ny);
if (bReadTraj && bAnalyze && (readmat[0].nx != nf))
gmx_fatal(FARGS,"Matrix size (%dx%d) does not match the number of "
"frames (%d)",readmat[0].nx,readmat[0].ny,nf);
nf = readmat[0].nx;
sfree(time);
time = readmat[0].axis_x;
time_invfac = time_invfactor();
for(i=0; i<nf; i++)
time[i] *= time_invfac;
rms = init_mat(readmat[0].nx,method == m_diagonalize);
convert_mat(&(readmat[0]),rms);
nlevels = readmat[0].nmap;
} else { /* !bReadMat */
rms = init_mat(nf,method == m_diagonalize);
nrms = (nf*(nf-1))/2;
if (!bRMSdist) {
fprintf(stderr,"Computing %dx%d RMS deviation matrix\n",nf,nf);
snew(x1,isize);
for(i1=0; (i1<nf); i1++) {
for(i2=i1+1; (i2<nf); i2++) {
for(i=0; i<isize; i++)
copy_rvec(xx[i1][i],x1[i]);
if (bFit)
do_fit(isize,mass,xx[i2],x1);
rmsd = rmsdev(isize,mass,xx[i2],x1);
set_mat_entry(rms,i1,i2,rmsd);
}
nrms -= (nf-i1-1);
fprintf(stderr,"\r# RMSD calculations left: %d ",nrms);
}
} else { /* bRMSdist */
fprintf(stderr,"Computing %dx%d RMS distance deviation matrix\n",nf,nf);
for(i1=0; (i1<nf); i1++) {
calc_dist(isize,xx[i1],d1);
for(i2=i1+1; (i2<nf); i2++) {
calc_dist(isize,xx[i2],d2);
set_mat_entry(rms,i1,i2,rms_dist(isize,d1,d2));
}
nrms -= (nf-i1-1);
fprintf(stderr,"\r# RMSD calculations left: %d ",nrms);
}
}
fprintf(stderr,"\n\n");
}
ffprintf2(stderr,log,buf,"The RMSD ranges from %g to %g nm\n",
rms->minrms,rms->maxrms);
ffprintf1(stderr,log,buf,"Average RMSD is %g\n",2*rms->sumrms/(nf*(nf-1)));
ffprintf1(stderr,log,buf,"Number of structures for matrix %d\n",nf);
ffprintf1(stderr,log,buf,"Energy of the matrix is %g nm\n",mat_energy(rms));
if (bUseRmsdCut && (rmsdcut < rms->minrms || rmsdcut > rms->maxrms) )
fprintf(stderr,"WARNING: rmsd cutoff %g is outside range of rmsd values "
"%g to %g\n",rmsdcut,rms->minrms,rms->maxrms);
if (bAnalyze && (rmsmin < rms->minrms) )
fprintf(stderr,"WARNING: rmsd minimum %g is below lowest rmsd value %g\n",
rmsmin,rms->minrms);
if (bAnalyze && (rmsmin > rmsdcut) )
fprintf(stderr,"WARNING: rmsd minimum %g is above rmsd cutoff %g\n",
rmsmin,rmsdcut);
/* Plot the rmsd distribution */
rmsd_distribution(opt2fn("-dist",NFILE,fnm),rms);
if (bBinary) {
for(i1=0; (i1 < nf); i1++)
for(i2=0; (i2 < nf); i2++)
if (rms->mat[i1][i2] < rmsdcut)
rms->mat[i1][i2] = 0;
else
rms->mat[i1][i2] = 1;
}
snew(clust.cl,nf);
switch (method) {
case m_linkage:
/* Now sort the matrix and write it out again */
gather(rms,rmsdcut,&clust);
break;
case m_diagonalize:
/* Do a diagonalization */
snew(eigval,nf);
snew(tmp,nf*nf);
memcpy(tmp,rms->mat[0],nf*nf*sizeof(real));
eigensolver(tmp,nf,0,nf,eigval,rms->mat[0]);
sfree(tmp);
fp = xvgropen(opt2fn("-ev",NFILE,fnm),"RMSD matrix Eigenvalues",
"Eigenvector index","Eigenvalues (nm\\S2\\N)");
for(i=0; (i<nf); i++)
fprintf(fp,"%10d %10g\n",i,eigval[i]);
ffclose(fp);
break;
case m_monte_carlo:
mc_optimize(log,rms,niter,&seed,kT);
swap_mat(rms);
reset_index(rms);
break;
case m_jarvis_patrick:
jarvis_patrick(rms->nn,rms->mat,M,P,bJP_RMSD ? rmsdcut : -1,&clust);
break;
case m_gromos:
gromos(rms->nn,rms->mat,rmsdcut,&clust);
break;
default:
gmx_fatal(FARGS,"DEATH HORROR unknown method \"%s\"",methodname[0]);
}
if (method == m_monte_carlo || method == m_diagonalize)
fprintf(stderr,"Energy of the matrix after clustering is %g nm\n",
mat_energy(rms));
if (bAnalyze) {
if (minstruct > 1) {
ncluster = plot_clusters(nf,rms->mat,&clust,nlevels,minstruct);
} else {
mark_clusters(nf,rms->mat,rms->maxrms,&clust);
}
init_t_atoms(&useatoms,isize,FALSE);
snew(usextps, isize);
useatoms.resname=top.atoms.resname;
for(i=0; i<isize; i++) {
useatoms.atomname[i]=top.atoms.atomname[index[i]];
useatoms.atom[i].resnr=top.atoms.atom[index[i]].resnr;
useatoms.nres=max(useatoms.nres,useatoms.atom[i].resnr+1);
copy_rvec(xtps[index[i]],usextps[i]);
}
useatoms.nr=isize;
analyze_clusters(nf,&clust,rms->mat,isize,&useatoms,usextps,mass,xx,time,
ifsize,fitidx,iosize,outidx,
bReadTraj?trx_out_fn:NULL,
opt2fn_null("-sz",NFILE,fnm),
opt2fn_null("-tr",NFILE,fnm),
opt2fn_null("-ntr",NFILE,fnm),
opt2fn_null("-clid",NFILE,fnm),
bAverage, write_ncl, write_nst, rmsmin, bFit, log,
rlo_bot,rhi_bot);
}
ffclose(log);
if (bBinary && !bAnalyze)
/* Make the clustering visible */
for(i2=0; (i2 < nf); i2++)
for(i1=i2+1; (i1 < nf); i1++)
if (rms->mat[i1][i2])
rms->mat[i1][i2] = rms->maxrms;
fp = opt2FILE("-o",NFILE,fnm,"w");
fprintf(stderr,"Writing rms distance/clustering matrix ");
if (bReadMat) {
write_xpm(fp,0,readmat[0].title,readmat[0].legend,readmat[0].label_x,
readmat[0].label_y,nf,nf,readmat[0].axis_x,readmat[0].axis_y,
rms->mat,0.0,rms->maxrms,rlo_top,rhi_top,&nlevels);
}
else {
sprintf(buf,"Time (%s)",time_unit());
sprintf(title,"RMS%sDeviation / Cluster Index",
bRMSdist ? " Distance " : " ");
if (minstruct > 1) {
write_xpm_split(fp,0,title,"RMSD (nm)",buf,buf,
nf,nf,time,time,rms->mat,0.0,rms->maxrms,&nlevels,
rlo_top,rhi_top,0.0,(real) ncluster,
&ncluster,TRUE,rlo_bot,rhi_bot);
} else {
write_xpm(fp,0,title,"RMSD (nm)",buf,buf,
nf,nf,time,time,rms->mat,0.0,rms->maxrms,
rlo_top,rhi_top,&nlevels);
}
}
fprintf(stderr,"\n");
ffclose(fp);
/* now show what we've done */
do_view(opt2fn("-o",NFILE,fnm),"-nxy");
do_view(opt2fn_null("-sz",NFILE,fnm),"-nxy");
if (method == m_diagonalize)
do_view(opt2fn_null("-ev",NFILE,fnm),"-nxy");
do_view(opt2fn("-dist",NFILE,fnm),"-nxy");
if (bAnalyze) {
do_view(opt2fn_null("-tr",NFILE,fnm),"-nxy");
do_view(opt2fn_null("-ntr",NFILE,fnm),"-nxy");
do_view(opt2fn_null("-clid",NFILE,fnm),"-nxy");
}
/* Thank the user for her patience */
thanx(stderr);
return 0;
}
int gmx_helix(int argc,char *argv[])
{
const char *desc[] = {
"g_helix computes all kind of helix properties. First, the peptide",
"is checked to find the longest helical part. This is determined by",
"Hydrogen bonds and Phi/Psi angles.",
"That bit is fitted",
"to an ideal helix around the Z-axis and centered around the origin.",
"Then the following properties are computed:[PAR]",
"[BB]1.[bb] Helix radius (file radius.xvg). This is merely the",
"RMS deviation in two dimensions for all Calpha atoms.",
"it is calced as sqrt((SUM i(x^2(i)+y^2(i)))/N), where N is the number",
"of backbone atoms. For an ideal helix the radius is 0.23 nm[BR]",
"[BB]2.[bb] Twist (file twist.xvg). The average helical angle per",
"residue is calculated. For alpha helix it is 100 degrees,",
"for 3-10 helices it will be smaller,",
"for 5-helices it will be larger.[BR]",
"[BB]3.[bb] Rise per residue (file rise.xvg). The helical rise per",
"residue is plotted as the difference in Z-coordinate between Ca",
"atoms. For an ideal helix this is 0.15 nm[BR]",
"[BB]4.[bb] Total helix length (file len-ahx.xvg). The total length",
"of the",
"helix in nm. This is simply the average rise (see above) times the",
"number of helical residues (see below).[BR]",
"[BB]5.[bb] Number of helical residues (file n-ahx.xvg). The title says",
"it all.[BR]",
"[BB]6.[bb] Helix Dipole, backbone only (file dip-ahx.xvg).[BR]",
"[BB]7.[bb] RMS deviation from ideal helix, calculated for the Calpha",
"atoms only (file rms-ahx.xvg).[BR]",
"[BB]8.[bb] Average Calpha-Calpha dihedral angle (file phi-ahx.xvg).[BR]",
"[BB]9.[bb] Average Phi and Psi angles (file phipsi.xvg).[BR]",
"[BB]10.[bb] Ellipticity at 222 nm according to [IT]Hirst and Brooks[it]",
"[PAR]"
};
static const char *ppp[efhNR+2] = {
NULL, "RAD", "TWIST", "RISE", "LEN", "NHX", "DIP", "RMS", "CPHI",
"RMSA", "PHI", "PSI", "HB3", "HB4", "HB5", "CD222", NULL
};
static gmx_bool bCheck=FALSE,bFit=TRUE,bDBG=FALSE,bEV=FALSE;
static int rStart=0,rEnd=0,r0=1;
t_pargs pa [] = {
{ "-r0", FALSE, etINT, {&r0},
"The first residue number in the sequence" },
{ "-q", FALSE, etBOOL,{&bCheck},
"Check at every step which part of the sequence is helical" },
{ "-F", FALSE, etBOOL,{&bFit},
"Toggle fit to a perfect helix" },
{ "-db", FALSE, etBOOL,{&bDBG},
"Print debug info" },
{ "-prop", FALSE, etENUM, {ppp},
"Select property to weight eigenvectors with. WARNING experimental stuff" },
{ "-ev", FALSE, etBOOL,{&bEV},
"Write a new 'trajectory' file for ED" },
{ "-ahxstart", FALSE, etINT, {&rStart},
"First residue in helix" },
{ "-ahxend", FALSE, etINT, {&rEnd},
"Last residue in helix" }
};
typedef struct {
FILE *fp,*fp2;
gmx_bool bfp2;
const char *filenm;
const char *title;
const char *xaxis;
const char *yaxis;
real val;
} t_xvgrfile;
t_xvgrfile xf[efhNR] = {
{ NULL, NULL, TRUE, "radius", "Helix radius", NULL, "r (nm)" , 0.0 },
{ NULL, NULL, TRUE, "twist", "Twist per residue", NULL, "Angle (deg)", 0.0 },
{ NULL, NULL, TRUE, "rise", "Rise per residue", NULL, "Rise (nm)", 0.0 },
{ NULL, NULL, FALSE, "len-ahx", "Length of the Helix", NULL, "Length (nm)", 0.0 },
{ NULL, NULL, FALSE, "dip-ahx", "Helix Backbone Dipole", NULL, "rq (nm e)", 0.0 },
{ NULL, NULL, TRUE, "rms-ahx", "RMS Deviation from Ideal Helix", NULL, "RMS (nm)", 0.0 },
{ NULL, NULL, FALSE, "rmsa-ahx","Average RMSD per Residue", "Residue", "RMS (nm)", 0.0 },
{ NULL, NULL,FALSE, "cd222", "Ellipticity at 222 nm", NULL, "nm", 0.0 },
{ NULL, NULL, TRUE, "pprms", "RMS Distance from \\8a\\4-helix", NULL, "deg" , 0.0 },
{ NULL, NULL, TRUE, "caphi", "Average Ca-Ca Dihedral", NULL, "\\8F\\4(deg)", 0.0 },
{ NULL, NULL, TRUE, "phi", "Average \\8F\\4 angles", NULL, "deg" , 0.0 },
{ NULL, NULL, TRUE, "psi", "Average \\8Y\\4 angles", NULL, "deg" , 0.0 },
{ NULL, NULL, TRUE, "hb3", "Average n-n+3 hbond length", NULL, "nm" , 0.0 },
{ NULL, NULL, TRUE, "hb4", "Average n-n+4 hbond length", NULL, "nm" , 0.0 },
{ NULL, NULL, TRUE, "hb5", "Average n-n+5 hbond length", NULL, "nm" , 0.0 },
{ NULL, NULL,FALSE, "JCaHa", "J-Coupling Values", "Residue", "Hz" , 0.0 },
{ NULL, NULL,FALSE, "helicity","Helicity per Residue", "Residue", "% of time" , 0.0 }
};
output_env_t oenv;
FILE *otrj;
char buf[54],prop[256];
t_trxstatus *status;
int natoms,nre,nres;
t_bb *bb;
int i,j,ai,m,nall,nbb,nca,teller,nSel=0;
atom_id *bbindex,*caindex,*allindex;
t_topology *top;
int ePBC;
rvec *x,*xref,*xav;
real t;
real rms,fac;
matrix box;
gmx_rmpbc_t gpbc=NULL;
gmx_bool bRange;
t_filenm fnm[] = {
{ efTPX, NULL, NULL, ffREAD },
{ efNDX, NULL, NULL, ffREAD },
{ efTRX, "-f", NULL, ffREAD },
{ efG87, "-to", NULL, ffOPTWR },
{ efSTO, "-cz", "zconf",ffWRITE },
{ efSTO, "-co", "waver",ffWRITE }
};
#define NFILE asize(fnm)
CopyRight(stderr,argv[0]);
parse_common_args(&argc,argv,PCA_CAN_VIEW | PCA_CAN_TIME | PCA_BE_NICE,
NFILE,fnm,asize(pa),pa,asize(desc),desc,0,NULL,&oenv);
bRange=(opt2parg_bSet("-ahxstart",asize(pa),pa) &&
opt2parg_bSet("-ahxend",asize(pa),pa));
top=read_top(ftp2fn(efTPX,NFILE,fnm),&ePBC);
natoms=read_first_x(oenv,&status,opt2fn("-f",NFILE,fnm),&t,&x,box);
if (opt2bSet("-to",NFILE,fnm)) {
otrj=opt2FILE("-to",NFILE,fnm,"w");
strcpy(prop,ppp[0]);
fprintf(otrj,"%s Weighted Trajectory: %d atoms, NO box\n",prop,natoms);
}
else
otrj=NULL;
if (natoms != top->atoms.nr)
gmx_fatal(FARGS,"Sorry can only run when the number of atoms in the run input file (%d) is equal to the number in the trajectory (%d)",
top->atoms.nr,natoms);
bb=mkbbind(ftp2fn(efNDX,NFILE,fnm),&nres,&nbb,r0,&nall,&allindex,
top->atoms.atomname,top->atoms.atom,top->atoms.resinfo);
snew(bbindex,natoms);
snew(caindex,nres);
fprintf(stderr,"nall=%d\n",nall);
/* Open output files, default x-axis is time */
for(i=0; (i<efhNR); i++) {
sprintf(buf,"%s.xvg",xf[i].filenm);
remove(buf);
xf[i].fp=xvgropen(buf,xf[i].title,
xf[i].xaxis ? xf[i].xaxis : "Time (ps)",
xf[i].yaxis,oenv);
if (xf[i].bfp2) {
sprintf(buf,"%s.out",xf[i].filenm);
remove(buf);
xf[i].fp2=ffopen(buf,"w");
}
}
/* Read reference frame from tpx file to compute helix length */
snew(xref,top->atoms.nr);
read_tpx(ftp2fn(efTPX,NFILE,fnm),
NULL,NULL,&natoms,xref,NULL,NULL,NULL);
calc_hxprops(nres,bb,xref,box);
do_start_end(nres,bb,xref,&nbb,bbindex,&nca,caindex,bRange,rStart,rEnd);
sfree(xref);
if (bDBG) {
fprintf(stderr,"nca=%d, nbb=%d\n",nca,nbb);
pr_bb(stdout,nres,bb);
}
gpbc = gmx_rmpbc_init(&top->idef,ePBC,natoms,box);
snew(xav,natoms);
teller=0;
do {
if ((teller++ % 10) == 0)
fprintf(stderr,"\rt=%.2f",t);
gmx_rmpbc(gpbc,natoms,box,x);
calc_hxprops(nres,bb,x,box);
if (bCheck)
do_start_end(nres,bb,x,&nbb,bbindex,&nca,caindex,FALSE,0,0);
if (nca >= 5) {
rms=fit_ahx(nres,bb,natoms,nall,allindex,x,nca,caindex,box,bFit);
if (teller == 1) {
write_sto_conf(opt2fn("-cz",NFILE,fnm),"Helix fitted to Z-Axis",
&(top->atoms),x,NULL,ePBC,box);
}
xf[efhRAD].val = radius(xf[efhRAD].fp2,nca,caindex,x);
xf[efhTWIST].val = twist(xf[efhTWIST].fp2,nca,caindex,x);
xf[efhRISE].val = rise(nca,caindex,x);
xf[efhLEN].val = ahx_len(nca,caindex,x,box);
xf[efhCD222].val = ellipticity(nres,bb);
xf[efhDIP].val = dip(nbb,bbindex,x,top->atoms.atom);
xf[efhRMS].val = rms;
xf[efhCPHI].val = ca_phi(nca,caindex,x,box);
xf[efhPPRMS].val = pprms(xf[efhPPRMS].fp2,nres,bb);
for(j=0; (j<=efhCPHI); j++)
fprintf(xf[j].fp, "%10g %10g\n",t,xf[j].val);
av_phipsi(xf[efhPHI].fp,xf[efhPSI].fp,xf[efhPHI].fp2,xf[efhPSI].fp2,
t,nres,bb);
av_hblen(xf[efhHB3].fp,xf[efhHB3].fp2,
xf[efhHB4].fp,xf[efhHB4].fp2,
xf[efhHB5].fp,xf[efhHB5].fp2,
t,nres,bb);
if (otrj)
dump_otrj(otrj,nall,allindex,x,xf[nSel].val,xav);
}
} while (read_next_x(oenv,status,&t,natoms,x,box));
fprintf(stderr,"\n");
gmx_rmpbc_done(gpbc);
close_trj(status);
if (otrj) {
ffclose(otrj);
fac=1.0/teller;
for(i=0; (i<nall); i++) {
ai=allindex[i];
for(m=0; (m<DIM); m++)
xav[ai][m]*=fac;
}
write_sto_conf_indexed(opt2fn("-co",NFILE,fnm),
"Weighted and Averaged conformation",
&(top->atoms),xav,NULL,ePBC,box,nall,allindex);
}
for(i=0; (i<nres); i++) {
if (bb[i].nrms > 0) {
fprintf(xf[efhRMSA].fp,"%10d %10g\n",r0+i,bb[i].rmsa/bb[i].nrms);
}
fprintf(xf[efhAHX].fp,"%10d %10g\n",r0+i,(bb[i].nhx*100.0)/(real )teller);
fprintf(xf[efhJCA].fp,"%10d %10g\n",
r0+i,140.3+(bb[i].jcaha/(double)teller));
}
for(i=0; (i<efhNR); i++) {
ffclose(xf[i].fp);
if (xf[i].bfp2)
ffclose(xf[i].fp2);
do_view(oenv,xf[i].filenm,"-nxy");
}
thanx(stderr);
return 0;
}
