// Action_NativeContacts::Init()
Action::RetType Action_NativeContacts::Init(ArgList& actionArgs, ActionInit& init, int debugIn)
{
# ifdef MPI
trajComm_ = init.TrajComm();
# endif
masterDSL_ = init.DslPtr();
debug_ = debugIn;
// Get Keywords
image_.InitImaging( !(actionArgs.hasKey("noimage")) );
double dist = actionArgs.getKeyDouble("distance", 7.0);
byResidue_ = actionArgs.hasKey("byresidue");
resoffset_ = actionArgs.getKeyInt("resoffset", 0) + 1;
if (resoffset_ < 1) {
mprinterr("Error: Residue offset must be >= 0\n");
return Action::ERR;
}
includeSolvent_ = actionArgs.hasKey("includesolvent");
series_ = actionArgs.hasKey("series");
saveNonNative_ = actionArgs.hasKey("savenonnative");
if (actionArgs.hasKey("skipnative"))
determineNativeContacts_ = false;
if (!determineNativeContacts_ && !saveNonNative_) {
mprintf("Warning: 'skipnative' specified; implies 'savenonnative'.\n");
saveNonNative_ = true;
}
# ifdef MPI
if (saveNonNative_) {
mprinterr("Error: Saving non-native contact data not yet supported for MPI\n");
return Action::ERR;
}
# endif
distance_ = dist * dist; // Square the cutoff
first_ = actionArgs.hasKey("first");
DataFile* outfile = init.DFL().AddDataFile( actionArgs.GetStringKey("out"), actionArgs );
Rseries_ = NO_RESSERIES;
if (series_) {
seriesout_ = init.DFL().AddDataFile(actionArgs.GetStringKey("seriesout"), actionArgs);
init.DSL().SetDataSetsPending( true );
if (saveNonNative_)
seriesNNout_ = init.DFL().AddDataFile(actionArgs.GetStringKey("seriesnnout"), actionArgs);
std::string rs_arg = actionArgs.GetStringKey("resseries");
if (!rs_arg.empty()) {
if (rs_arg == "present")
Rseries_ = RES_PRESENT;
else if (rs_arg == "sum")
Rseries_ = RES_SUM;
else {
mprinterr("Error: '%s' is not a valid 'resseries' keyword.\n", rs_arg.c_str());
return Action::ERR;
}
seriesRout_ = init.DFL().AddDataFile(actionArgs.GetStringKey("resseriesout"), actionArgs);
}
} else {
if (KeywordError(actionArgs,"seriesout")) return Action::ERR;
if (KeywordError(actionArgs,"seriesnnout")) return Action::ERR;
if (KeywordError(actionArgs,"resseries")) return Action::ERR;
if (KeywordError(actionArgs,"resseriesout")) return Action::ERR;
}
cfile_ = init.DFL().AddCpptrajFile(actionArgs.GetStringKey("writecontacts"), "Native Contacts",
DataFileList::TEXT, true);
pfile_ = init.DFL().AddCpptrajFile(actionArgs.GetStringKey("contactpdb"), "Contact PDB",
DataFileList::PDB);
if (saveNonNative_)
nfile_ = init.DFL().AddCpptrajFile(actionArgs.GetStringKey("nncontactpdb"),
"Non-native Contact PDB", DataFileList::PDB);
rfile_ = init.DFL().AddCpptrajFile(actionArgs.GetStringKey("resout"), "Contact Res Pairs",
DataFileList::TEXT, true);
if (cfile_ == 0 || rfile_ == 0) return Action::ERR;
pdbcut_ = (float)actionArgs.getKeyDouble("pdbcut", -1.0);
usepdbcut_ = (pdbcut_ > -1.0);
// Get reference for native contacts. Do this even if we wont be
// determining native contacts in order to set up contact lists.
ReferenceFrame REF = init.DSL().GetReferenceFrame( actionArgs );
if (!first_) {
if (REF.error()) return Action::ERR;
if (REF.empty()) {
mprintf("Warning: No reference structure specified. Defaulting to first.\n");
first_ = true;
}
} else {
if (!REF.empty()) {
mprinterr("Error: Must only specify 'first' or a reference structure, not both.\n");
return Action::ERR;
}
}
// Create data sets
std::string name = actionArgs.GetStringKey("name");
if (name.empty())
name = init.DSL().GenerateDefaultName("Contacts");
numnative_ = init.DSL().AddSet(DataSet::INTEGER, MetaData(name, "native"));
nonnative_ = init.DSL().AddSet(DataSet::INTEGER, MetaData(name, "nonnative"));
if (outfile != 0) {
outfile->AddDataSet(numnative_);
outfile->AddDataSet(nonnative_);
}
if (numnative_ == 0 || nonnative_ == 0) return Action::ERR;
if (actionArgs.hasKey("mindist")) {
mindist_ = init.DSL().AddSet(DataSet::DOUBLE, MetaData(name, "mindist"));
if (mindist_ == 0) return Action::ERR;
if (outfile != 0) outfile->AddDataSet(mindist_);
}
if (actionArgs.hasKey("maxdist")) {
maxdist_ = init.DSL().AddSet(DataSet::DOUBLE, MetaData(name, "maxdist"));
if (maxdist_ == 0) return Action::ERR;
if (outfile != 0) outfile->AddDataSet(maxdist_);
}
DataFile *natmapfile = 0, *nonmapfile = 0;
if (actionArgs.hasKey("map")) {
nativeMap_ = (DataSet_MatrixDbl*)init.DSL().AddSet(DataSet::MATRIX_DBL, MetaData(name, "nativemap"));
if (nativeMap_ == 0) return Action::ERR;
nonnatMap_ = (DataSet_MatrixDbl*)init.DSL().AddSet(DataSet::MATRIX_DBL, MetaData(name, "nonnatmap"));
if (nonnatMap_ == 0) return Action::ERR;
FileName mapFilename;
mapFilename.SetFileName( actionArgs.GetStringKey("mapout") );
if (!mapFilename.empty()) {
natmapfile = init.DFL().AddDataFile(mapFilename.PrependFileName("native."));
if (natmapfile != 0) natmapfile->AddDataSet(nativeMap_);
nonmapfile = init.DFL().AddDataFile(mapFilename.PrependFileName("nonnative."));
if (nonmapfile != 0) nonmapfile->AddDataSet(nonnatMap_);
}
}
// Get Masks
if (Mask1_.SetMaskString( actionArgs.GetMaskNext() )) return Action::ERR;
std::string mask2 = actionArgs.GetMaskNext();
if (!mask2.empty()) {
if (Mask2_.SetMaskString( mask2 )) return Action::ERR;
}
mprintf(" NATIVECONTACTS: Mask1='%s'", Mask1_.MaskString());
if (Mask2_.MaskStringSet())
mprintf(" Mask2='%s'", Mask2_.MaskString());
if (determineNativeContacts_) {
mprintf(", native contacts set up based on");
if (first_)
mprintf(" first frame.\n");
else
mprintf("'%s'.\n", REF.refName());
} else {
mprintf(", skipping native contacts set up.\n");
}
if (byResidue_) {
mprintf("\tContacts will be ignored for residues spaced < %i apart.\n", resoffset_);
if (nativeMap_ != 0)
mprintf("\tMap will be printed by residue.\n");
}
if (saveNonNative_)
mprintf("\tSaving non-native contacts as well (may use a lot of memory).\n");
mprintf("\tDistance cutoff is %g Angstroms,", sqrt(distance_));
if (!image_.UseImage())
mprintf(" imaging is off.\n");
else
mprintf(" imaging is on.\n");
if (includeSolvent_)
mprintf("\tMask selection will including solvent.\n");
else
mprintf("\tMask selection will not include solvent.\n");
mprintf("\tData set name: %s\n", name.c_str());
if (maxdist_ != 0)
mprintf("\tSaving maximum observed distances in set '%s'\n", maxdist_->legend());
if (mindist_ != 0)
mprintf("\tSaving minimum observed distances in set '%s'\n", mindist_->legend());
if (outfile != 0)
mprintf("\tOutput to '%s'\n", outfile->DataFilename().full());
mprintf("\tContact stats will be written to '%s'\n", cfile_->Filename().full());
mprintf("\tContact res pairs will be written to '%s'\n", rfile_->Filename().full());
if (pfile_ != 0) {
mprintf("\tContact PDB will be written to '%s'\n", pfile_->Filename().full());
if (usepdbcut_) mprintf("\tOnly atoms with values > %g will be written to PDB.\n", pdbcut_);
}
if (nfile_ != 0) {
mprintf("\tNon-native contact PDB will be written to '%s'\n", nfile_->Filename().full());
if (usepdbcut_) mprintf("\tOnly atoms with values > %g will be written to PDB.\n", pdbcut_);
}
if (nativeMap_ != 0) {
mprintf("\tNative contacts map will be saved as set '%s'\n"
"\tNon-native contacts map will be saved as set '%s'\n",
nativeMap_->legend(), nonnatMap_->legend());
if (natmapfile!=0) mprintf("\tNative map output to '%s'\n",natmapfile->DataFilename().full());
if (nonmapfile!=0) mprintf("\tNative map output to '%s'\n",nonmapfile->DataFilename().full());
}
if (series_) {
mprintf("\tSaving native contact time series %s[NC].\n", name.c_str());
if (seriesout_ != 0) mprintf("\tWriting native contact time series to %s\n",
seriesout_->DataFilename().full());
if (saveNonNative_) {
mprintf("\tSaving non-native contact time series %s[NN]\n", name.c_str());
if (seriesNNout_ != 0) mprintf("\tWriting non-native contact time series to %s\n",
seriesNNout_->DataFilename().full());
}
if (Rseries_ != NO_RESSERIES) {
if (Rseries_ == RES_PRESENT)
mprintf("\tResidue contact time series will reflect presence of individual contacts.\n");
else if (Rseries_ == RES_SUM)
mprintf("\tResidue contact time series will reflect sum of individual contacts.\n");
if (seriesRout_ != 0) mprintf("\tWriting residue contact time series to %s\n",
seriesRout_->DataFilename().full());
}
}
// Set up reference if necessary.
if (!first_) {
// Set up imaging info for ref parm
image_.SetupImaging( REF.CoordsInfo().TrajBox().Type() );
if (image_.ImageType() == NONORTHO)
REF.Coord().BoxCrd().ToRecip(ucell_, recip_);
if (DetermineNativeContacts( REF.Parm(), REF.Coord() )) return Action::ERR;
}
return Action::OK;
}
// Action_Spam::init()
Action::RetType Action_Spam::Init(ArgList& actionArgs, TopologyList* PFL, DataSetList* DSL, DataFileList* DFL, int debugIn)
{
// Always use imaged distances
InitImaging(true);
// This is needed everywhere in this function scope
FileName filename;
// See if we're doing pure water. If so, we don't need a peak file
purewater_ = actionArgs.hasKey("purewater");
if (purewater_) {
// We still need the cutoff
double cut = actionArgs.getKeyDouble("cut", 12.0);
cut2_ = cut * cut;
doublecut_ = 2 * cut;
onecut2_ = 1 / cut2_;
// See if we write to a data file
datafile_ = actionArgs.GetStringKey("out");
// Generate the data set name, and hold onto the master data set list
std::string ds_name = actionArgs.GetStringKey("name");
if (ds_name.empty())
ds_name = myDSL_.GenerateDefaultName("SPAM");
// We only have one data set averaging over every water. Add it here
myDSL_.AddSet(DataSet::DOUBLE, ds_name, NULL);
solvname_ = actionArgs.GetStringKey("solv");
if (solvname_.empty())
solvname_ = std::string("WAT");
}else {
// Get the file name with the peaks defined in it
filename.SetFileName( actionArgs.GetStringNext() );
if (filename.empty() || !File::Exists(filename)) {
mprinterr("Spam: Error: Peak file [%s] does not exist!\n", filename.full());
return Action::ERR;
}
// Get the remaining optional arguments
solvname_ = actionArgs.GetStringKey("solv");
if (solvname_.empty())
solvname_ = std::string("WAT");
reorder_ = actionArgs.hasKey("reorder");
bulk_ = actionArgs.getKeyDouble("bulk", 0.0);
double cut = actionArgs.getKeyDouble("cut", 12.0);
cut2_ = cut * cut;
doublecut_ = 2 * cut;
onecut2_ = 1 / cut2_;
std::string infoname = actionArgs.GetStringKey("info");
if (infoname.empty())
infoname = std::string("spam.info");
infofile_ = DFL->AddCpptrajFile(infoname, "SPAM info");
if (infofile_ == 0) return Action::ERR;
// The default maskstr is the Oxygen atom of the solvent
summaryfile_ = actionArgs.GetStringKey("summary");
// Divide site size by 2 to make it half the edge length (or radius)
site_size_ = actionArgs.getKeyDouble("site_size", 2.5) / 2.0;
sphere_ = actionArgs.hasKey("sphere");
// If it's a sphere, square the radius to compare with
if (sphere_)
site_size_ *= site_size_;
datafile_ = actionArgs.GetStringKey("out");
std::string ds_name = actionArgs.GetStringKey("name");
if (ds_name.empty())
ds_name = myDSL_.GenerateDefaultName("SPAM");
// Parse through the peaks file and extract the peaks
CpptrajFile peakfile;
if (peakfile.OpenRead(filename)) {
mprinterr("SPAM: Error: Could not open %s for reading!\n", filename.full());
return Action::ERR;
}
std::string line = peakfile.GetLine();
int npeaks = 0;
while (!line.empty()) {
if (sscanf(line.c_str(), "%d", &npeaks) != 1) {
line = peakfile.GetLine();
continue;
}
line = peakfile.GetLine();
break;
}
while (!line.empty()) {
double x, y, z, dens;
if (sscanf(line.c_str(), "C %lg %lg %lg %lg", &x, &y, &z, &dens) != 4) {
line = peakfile.GetLine();
continue;
}
line = peakfile.GetLine();
peaks_.push_back(Vec3(x, y, z));
}
peakfile.CloseFile();
// Check that our initial number of peaks matches our parsed peaks. Warn
// otherwise
if (npeaks != (int)peaks_.size())
mprinterr("SPAM: Warning: %s claims to have %d peaks, but really has %d!\n",
filename.full(), npeaks, peaks_.size());
// Now add all of the data sets
MetaData md(ds_name);
for (int i = 0; i < (int)peaks_.size(); i++) {
md.SetAspect( integerToString(i+1) ); // TODO: Should this be Idx?
if (myDSL_.AddSet(DataSet::DOUBLE, md) == 0) return Action::ERR;
// Add a new list of integers to keep track of omitted frames
std::vector<int> vec;
peakFrameData_.push_back(vec);
}
}
// Print info now
if (purewater_) {
mprintf("SPAM: Calculating bulk value for pure solvent\n");
if (!datafile_.empty())
mprintf("SPAM: Printing solvent energies to %s\n", datafile_.c_str());
mprintf("SPAM: Using a %.2f Angstrom non-bonded cutoff with shifted EEL.\n",
sqrt(cut2_));
if (reorder_)
mprintf("SPAM: Warning: Re-ordering makes no sense for pure solvent.\n");
if (!summaryfile_.empty())
mprintf("SPAM: Printing solvent SPAM summary to %s\n", summaryfile_.c_str());
}else {
mprintf("SPAM: Solvent [%s] density peaks taken from %s.\n",
solvname_.c_str(), filename.base());
mprintf("SPAM: %d density peaks will be analyzed from %s.\n",
peaks_.size(), filename.base());
mprintf("SPAM: Occupation information printed to %s.\n", infofile_->Filename().full());
mprintf("SPAM: Sites are ");
if (sphere_)
mprintf("spheres with diameter %.3lf\n", site_size_);
else
mprintf("boxes with edge length %.3lf\n", site_size_);
if (reorder_) {
mprintf("SPAM: Re-ordering trajectory so each site always has ");
mprintf("the same water molecule.\n");
}
if (summaryfile_.empty() && datafile_.empty()) {
if (!reorder_) {
mprinterr("SPAM: Error: Not re-ordering trajectory or calculating energies. ");
mprinterr("Nothing to do!\n");
return Action::ERR;
}
mprintf("SPAM: Not calculating any SPAM energies\n");
}else {
mprintf("SPAM: Using a non-bonded cutoff of %.2lf Ang. with a EEL shifting function.\n",
sqrt(cut2_));
mprintf("SPAM: Bulk solvent SPAM energy taken as %.3lf kcal/mol\n", bulk_);
}
}
mprintf("#Citation: Cui, G.; Swails, J.M.; Manas, E.S.; \"SPAM: A Simple Approach\n"
"# for Profiling Bound Water Molecules\"\n"
"# J. Chem. Theory Comput., 2013, 9 (12), pp 5539–5549.\n");
return Action::OK;
}
