void CheckValidDipeptide(OBConversion &conv,
const string &test,
unsigned int testCount)
{
OBMol mol;
OBResidue *res;
ostringstream os;
mol.Clear();
conv.ReadString(&mol, test);
chainsparser.PerceiveChains(mol);
if (mol.NumResidues() != 2) {
os << "not ok " << testCount << " # expected 2 residues, but found "
<< mol.NumResidues() << '\n';
os << "# ";
FOR_RESIDUES_OF_MOL(res, mol)
os << res->GetName() << " ";
os << endl;
BOOST_CHECK_MESSAGE( 0, os.str().c_str() );
} else {
res = mol.GetResidue(0);
BOOST_CHECK_MESSAGE( res, "Get first AA from dipeptide" );
res = mol.GetResidue(1);
BOOST_CHECK_MESSAGE( res, "Get second AA from dipeptide" );
}
}
bool WriteGromos96(ostream &ofs,OBMol &mol,double fac)
{
char type_name[10];
char res_name[10],padded_name[10];
char buffer[BUFF_SIZE];
int res_num;
sprintf(buffer,"#GENERATED BY OPEN BABEL %s",BABEL_VERSION);
ofs << buffer << endl;
/* GROMOS wants a TITLE block, so let's write one*/
sprintf(buffer,"TITLE\n%s\nEND",mol.GetTitle());
ofs << buffer << endl;
ofs << "POSITION" << endl;
OBAtom *atom;
OBResidue *res;
vector<OBNodeBase*>::iterator i;
for(atom = mol.BeginAtom(i);atom;atom = mol.NextAtom(i))
{
if (res = atom->GetResidue())
{
strcpy(res_name,(char*)res->GetName().c_str());
strcpy(type_name,(char*)res->GetAtomID(atom).c_str());
res_num = res->GetNum();
}
else
{
strcpy(type_name,etab.GetSymbol(atom->GetAtomicNum()));
strcpy(res_name,"UNK");
sprintf(padded_name,"%2s",type_name);
strcpy(type_name,padded_name);
res_num = 1;
}
sprintf(buffer,"%5d %5s %5s %6d %15.5f %15.5f %15.5f",
res_num,res_name,type_name,atom->GetIdx(),
atom->x()*fac,atom->y()*fac,atom->z()*fac);
ofs << buffer << endl;
if (!(atom->GetIdx()%10))
{
sprintf(buffer,"# %d",atom->GetIdx());
ofs << buffer << endl;
}
}
ofs << "END" << endl;
return(true);
}
void CheckInvalidResidue(OBConversion &conv,
const string &test,
unsigned int testCount)
{
OBMol mol;
mol.Clear();
conv.ReadString(&mol, test);
chainsparser.PerceiveChains(mol);
if (mol.NumResidues() != 0) {
OBResidue *res = mol.GetResidue(0);
if (res->GetName() == "LIG") { // ligand, not residue
cout << "ok " << testCount << " # found ligand, not residue "
<< test << '\n';
} else {
cout << "not ok " << testCount << " # expected 0 residues, found "
<< mol.NumResidues() << '\n';
cout << "# " << res->GetName() << endl;
}
} else
cout << "ok " << testCount << " # correctly rejected " << test << '\n';
}
void CheckInvalidResidue(OBConversion &conv,
const string &test,
unsigned int testCount)
{
OBMol mol;
ostringstream os;
mol.Clear();
conv.ReadString(&mol, test);
chainsparser.PerceiveChains(mol);
if (mol.NumResidues() != 0) {
OBResidue *res = mol.GetResidue(0);
if (res->GetName() == "LIG") { // ligand, not residue
BOOST_CHECK( 1 );
} else {
os << "not ok " << testCount << " # expected 0 residues, found "
<< mol.NumResidues() << '\n';
os << "# " << res->GetName() << endl;
BOOST_CHECK_MESSAGE( 0, os.str().c_str() );
}
} else
BOOST_CHECK( 1 );
}
void CheckValidDipeptide(OBConversion &conv,
const string &test,
unsigned int testCount)
{
OBMol mol;
mol.Clear();
conv.ReadString(&mol, test);
chainsparser.PerceiveChains(mol);
if (mol.NumResidues() != 2) {
cout << "not ok " << testCount << " # expected 2 residues, but found "
<< mol.NumResidues() << '\n';
cout << "# ";
FOR_RESIDUES_OF_MOL(res, mol)
cout << res->GetName() << " ";
cout << endl;
} else {
OBResidue *res;
res = mol.GetResidue(0);
cout << "ok " << testCount << " # " << res->GetName();
res = mol.GetResidue(1);
cout << " " << res->GetName() << '\n';
}
}
bool MOL2Format::ReadMolecule(OBBase* pOb, OBConversion* pConv)
{
OBMol* pmol = pOb->CastAndClear<OBMol>();
if(pmol==NULL)
return false;
//Define some references so we can use the old parameter names
istream &ifs = *pConv->GetInStream();
OBMol &mol = *pmol;
//Old code follows...
bool foundAtomLine = false;
char buffer[BUFF_SIZE];
char *comment = NULL;
string str,str1;
vector<string> vstr;
int len;
mol.BeginModify();
for (;;)
{
if (!ifs.getline(buffer,BUFF_SIZE))
return(false);
if (EQn(buffer,"@<TRIPOS>MOLECULE",17))
break;
}
int lcount;
int natoms,nbonds;
for (lcount=0;;lcount++)
{
if (!ifs.getline(buffer,BUFF_SIZE))
return(false);
if (EQn(buffer,"@<TRIPOS>ATOM",13))
{
foundAtomLine = true;
break;
}
if (lcount == 0)
{
tokenize(vstr,buffer);
if (!vstr.empty())
mol.SetTitle(buffer);
}
else if (lcount == 1)
sscanf(buffer,"%d%d",&natoms,&nbonds);
else if (lcount == 4) //energy
{
tokenize(vstr,buffer);
if (!vstr.empty() && vstr.size() == 3)
if (vstr[0] == "Energy")
mol.SetEnergy(atof(vstr[2].c_str()));
}
else if (lcount == 5) //comment
{
if ( buffer[0] )
{
len = (int) strlen(buffer)+1;
// TODO allow better multi-line comments
// which don't allow ill-formed data to consume memory
// Thanks to Andrew Dalke for the pointer
if (comment != NULL)
delete [] comment;
comment = new char [len];
memcpy(comment,buffer,len);
}
}
}
if (!foundAtomLine)
{
mol.EndModify();
mol.Clear();
obErrorLog.ThrowError(__FUNCTION__, "Unable to read Mol2 format file. No atoms found.", obWarning);
return(false);
}
mol.ReserveAtoms(natoms);
int i;
vector3 v;
OBAtom atom;
bool hasPartialCharges=false;
double x,y,z,pcharge;
char temp_type[BUFF_SIZE], resname[BUFF_SIZE], atmid[BUFF_SIZE];
int elemno, resnum = -1;
ttab.SetFromType("SYB");
for (i = 0;i < natoms;i++)
{
if (!ifs.getline(buffer,BUFF_SIZE))
return(false);
sscanf(buffer," %*s %1024s %lf %lf %lf %1024s %d %1024s %lf",
atmid, &x,&y,&z, temp_type, &resnum, resname, &pcharge);
atom.SetVector(x, y, z);
// Handle "CL" and "BR" and other mis-typed atoms
str = temp_type;
if (strncmp(temp_type, "CL", 2) == 0) {
str = "Cl";
} else if (strncmp(temp_type,"BR",2) == 0) {
str = "Br";
} else if (strncmp(temp_type,"S.o2", 4) == 02) {
str = "S.O2";
} else if (strncmp(temp_type,"S.o", 3) == 0) {
str = "S.O";
} else if (strncmp(temp_type,"SI", 2) == 0) {
str = "Si";
// The following cases are entries which are not in openbabel/data/types.txt
// and should probably be added there
} else if (strncmp(temp_type,"S.1", 3) == 0) {
str = "S.2"; // no idea what the best type might be here
} else if (strncmp(temp_type,"P.", 2) == 0) {
str = "P.3";
} else if (strncasecmp(temp_type,"Ti.", 3) == 0) { // e.g. Ti.th
str = "Ti";
} else if (strncasecmp(temp_type,"Ru.", 3) == 0) { // e.g. Ru.oh
str = "Ru";
}
ttab.SetToType("ATN");
ttab.Translate(str1,str);
elemno = atoi(str1.c_str());
ttab.SetToType("IDX");
// We might have missed some SI or FE type things above, so here's
// another check
if( !elemno && isupper(temp_type[1]) )
{
temp_type[1] = (char)tolower(temp_type[1]);
str = temp_type;
ttab.Translate(str1,str);
elemno = atoi(str1.c_str());
}
// One last check if there isn't a period in the type,
// it's a malformed atom type, but it may be the element symbol
// GaussView does this (PR#1739905)
if ( !elemno ) {
obErrorLog.ThrowError(__FUNCTION__, "This Mol2 file is non-standard. Cannot interpret atom types correctly, instead attempting to interpret as elements instead.", obWarning);
string::size_type dotPos = str.find('.');
if (dotPos == string::npos) {
elemno = etab.GetAtomicNum(str.c_str());
}
}
atom.SetAtomicNum(elemno);
ttab.SetToType("INT");
ttab.Translate(str1,str);
atom.SetType(str1);
atom.SetPartialCharge(pcharge);
if (!mol.AddAtom(atom))
return(false);
if (!IsNearZero(pcharge))
hasPartialCharges = true;
// Add residue information if it exists
if (resnum != -1 && resnum != 0 &&
strlen(resname) != 0 && strncmp(resname,"<1>", 3) != 0)
{
OBResidue *res = (mol.NumResidues() > 0) ?
mol.GetResidue(mol.NumResidues()-1) : NULL;
if (res == NULL || res->GetName() != resname ||
static_cast<int>(res->GetNum()) != resnum)
{
vector<OBResidue*>::iterator ri;
for (res = mol.BeginResidue(ri) ; res ; res = mol.NextResidue(ri))
if (res->GetName() == resname &&
static_cast<int>(res->GetNum())
== resnum)
break;
if (res == NULL)
{
res = mol.NewResidue();
res->SetName(resname);
res->SetNum(resnum);
}
}
OBAtom *atomPtr = mol.GetAtom(mol.NumAtoms());
res->AddAtom(atomPtr);
res->SetAtomID(atomPtr, atmid);
} // end adding residue info
}
for (;;)
{
if (!ifs.getline(buffer,BUFF_SIZE))
return(false);
str = buffer;
if (!strncmp(buffer,"@<TRIPOS>BOND",13))
break;
}
int start,end,order;
for (i = 0; i < nbonds; i++)
{
if (!ifs.getline(buffer,BUFF_SIZE))
return(false);
sscanf(buffer,"%*d %d %d %1024s",&start,&end,temp_type);
str = temp_type;
order = 1;
if (str == "ar" || str == "AR" || str == "Ar")
order = 5;
else if (str == "AM" || str == "am" || str == "Am")
order = 1;
else
order = atoi(str.c_str());
mol.AddBond(start,end,order);
}
// update neighbour bonds information for each atom.
vector<OBAtom*>::iterator apos;
vector<OBBond*>::iterator bpos;
OBAtom* patom;
OBBond* pbond;
for (patom = mol.BeginAtom(apos); patom; patom = mol.NextAtom(apos))
{
patom->ClearBond();
for (pbond = mol.BeginBond(bpos); pbond; pbond = mol.NextBond(bpos))
{
if (patom == pbond->GetBeginAtom() || patom == pbond->GetEndAtom())
{
patom->AddBond(pbond);
}
}
}
// Suggestion by Liu Zhiguo 2008-01-26
// Mol2 files define atom types -- there is no need to re-perceive
mol.SetAtomTypesPerceived();
mol.EndModify();
//must add generic data after end modify - otherwise it will be blown away
if (comment)
{
OBCommentData *cd = new OBCommentData;
cd->SetData(comment);
cd->SetOrigin(fileformatInput);
mol.SetData(cd);
delete [] comment;
comment = NULL;
}
if (hasPartialCharges)
mol.SetPartialChargesPerceived();
// continue untill EOF or untill next molecule record
streampos pos;
for(;;)
{
pos = ifs.tellg();
if (!ifs.getline(buffer,BUFF_SIZE))
break;
if (EQn(buffer,"@<TRIPOS>MOLECULE",17))
break;
}
ifs.seekg(pos); // go back to the end of the molecule
return(true);
}
bool MOL2Format::WriteMolecule(OBBase* pOb, OBConversion* pConv)
{
OBMol* pmol = dynamic_cast<OBMol*>(pOb);
if(pmol==NULL)
return false;
//Define some references so we can use the old parameter names
ostream &ofs = *pConv->GetOutStream();
OBMol &mol = *pmol;
bool ligandsOnly = pConv->IsOption("l", OBConversion::OUTOPTIONS)!=NULL;
//The old code follows....
string str,str1;
char buffer[BUFF_SIZE],label[BUFF_SIZE];
char rnum[BUFF_SIZE],rlabel[BUFF_SIZE];
ofs << "@<TRIPOS>MOLECULE" << endl;
str = mol.GetTitle();
if (str.empty())
ofs << "*****" << endl;
else
ofs << str << endl;
snprintf(buffer, BUFF_SIZE," %d %d 0 0 0", mol.NumAtoms(),mol.NumBonds());
ofs << buffer << endl;
ofs << "SMALL" << endl;
OBPairData *dp = (OBPairData*)mol.GetData("PartialCharges");
if (dp != NULL) {
// Tripos spec says:
// NO_CHARGES, DEL_RE, GASTEIGER, GAST_HUCK, HUCKEL, PULLMAN,
// GAUSS80_CHARGES, AMPAC_CHARGES, MULLIKEN_CHARGES, DICT_ CHARGES,
// MMFF94_CHARGES, USER_CHARGES
if (dp->GetValue() == "Mulliken")
ofs << "MULLIKEN_CHARGES" << endl;
else // should pick from the Tripos types
ofs << "GASTEIGER" << endl;
}
else { // No idea what these charges are... all our code sets "PartialCharges"
ofs << "GASTEIGER" << endl;
}
ofs << "Energy = " << mol.GetEnergy() << endl;
if (mol.HasData(OBGenericDataType::CommentData))
{
OBCommentData *cd = (OBCommentData*)mol.GetData(OBGenericDataType::CommentData);
ofs << cd->GetData();
}
ofs << endl;
ofs << "@<TRIPOS>ATOM" << endl;
OBAtom *atom;
OBResidue *res;
vector<OBAtom*>::iterator i;
vector<int> labelcount;
labelcount.resize( etab.GetNumberOfElements() );
ttab.SetFromType("INT");
ttab.SetToType("SYB");
for (atom = mol.BeginAtom(i);atom;atom = mol.NextAtom(i))
{
//
// Use sequentially numbered atom names if no residues
//
snprintf(label,BUFF_SIZE, "%s%d",
etab.GetSymbol(atom->GetAtomicNum()),
++labelcount[atom->GetAtomicNum()]);
strcpy(rlabel,"<1>");
strcpy(rnum,"1");
str = atom->GetType();
ttab.Translate(str1,str);
//
// Use original atom names if there are residues
//
if (!ligandsOnly && (res = atom->GetResidue()) )
{
// use original atom names defined by residue
snprintf(label,BUFF_SIZE,"%s",(char*)res->GetAtomID(atom).c_str());
// make sure that residue name includes its number
snprintf(rlabel,BUFF_SIZE,"%s%d",res->GetName().c_str(), res->GetNum());
snprintf(rnum,BUFF_SIZE,"%d",res->GetNum());
}
snprintf(buffer,BUFF_SIZE,"%7d%1s%-6s%12.4f%10.4f%10.4f%1s%-5s%4s%1s %-8s%10.4f",
atom->GetIdx(),"",label,
atom->GetX(),atom->GetY(),atom->GetZ(),
"",str1.c_str(),
rnum,"",rlabel,
atom->GetPartialCharge());
ofs << buffer << endl;
}
ofs << "@<TRIPOS>BOND" << endl;
OBBond *bond;
vector<OBBond*>::iterator j;
OBSmartsPattern pat;
string s1, s2;
for (bond = mol.BeginBond(j);bond;bond = mol.NextBond(j))
{
s1 = bond->GetBeginAtom()->GetType();
s2 = bond->GetEndAtom()->GetType();
if (bond->IsAromatic() || s1 == "O.co2" || s2 == "O.co2")
strcpy(label,"ar");
else if (bond->IsAmide())
strcpy(label,"am");
else
snprintf(label,BUFF_SIZE,"%d",bond->GetBO());
snprintf(buffer, BUFF_SIZE,"%6d%6d%6d%3s%2s",
bond->GetIdx()+1,bond->GetBeginAtomIdx(),bond->GetEndAtomIdx(),
"",label);
ofs << buffer << endl;
}
// NO trailing blank line (PR#1868929).
// ofs << endl;
return(true);
}
bool GROMOS96Format::WriteMolecule(OBBase* pOb, OBConversion* pConv)
{
OBMol* pmol = dynamic_cast<OBMol*>(pOb);
if(pmol==NULL)
return false;
//Define some references so we can use the old parameter names
ostream &ofs = *pConv->GetOutStream();
OBMol &mol = *pmol;
double fac = pConv->IsOption("n") ? 0.1 : 1.0; //new framework
char type_name[16];
char res_name[16];
char buffer[BUFF_SIZE];
string res_num;
snprintf(buffer, BUFF_SIZE, "#GENERATED BY OPEN BABEL %s\n",BABEL_VERSION);
ofs << buffer;
/* GROMOS wants a TITLE block, so let's write one*/
ofs << "TITLE\n" << mol.GetTitle() << "\nEND\n";
ofs << "POSITION\n";
OBAtom *atom;
OBResidue *res;
vector<OBAtom*>::iterator i;
for(atom = mol.BeginAtom(i);atom;atom = mol.NextAtom(i))
{
if ( (res = atom->GetResidue()) )
{
// 16 = sizeof(res_name) and sizeof(type_name)
strncpy(res_name,(char*)res->GetName().c_str(), 16);
res_name[15] = '\0';
strncpy(type_name,(char*)res->GetAtomID(atom).c_str(), 16);
type_name[15] = '\0';
res_num = res->GetNumString();
}
else
{
strncpy(type_name,OBElements::GetSymbol(atom->GetAtomicNum()), 16);
strcpy(res_name,"UNK");
res_num = "1";
}
snprintf(buffer, BUFF_SIZE, "%5s %5s %5s %6d %15.5f %15.5f %15.5f\n",
res_num.c_str(),res_name,type_name,atom->GetIdx(),
atom->x()*fac,atom->y()*fac,atom->z()*fac);
ofs << buffer;
if (!(atom->GetIdx()%10))
{
snprintf(buffer, BUFF_SIZE, "# %d\n",atom->GetIdx());
ofs << buffer;
}
}
ofs << "END\n";
return(true);
}
static bool parseAtomRecord(char *buffer, OBMol &mol,int /*chainNum*/)
/* ATOMFORMAT "(i5,1x,a4,a1,a3,1x,a1,i4,a1,3x,3f8.3,2f6.2,a2,a2)" */
{
string sbuf = &buffer[6];
if (sbuf.size() < 48)
return(false);
bool hetatm = (EQn(buffer,"HETATM",6)) ? true : false;
bool elementFound = false; // true if correct element found in col 77-78
/* serial number */
string serno = sbuf.substr(0,5);
/* atom name */
string atmid = sbuf.substr(6,4);
/* chain */
char chain = sbuf.substr(15,1)[0];
/* element */
string element = " ";
if (sbuf.size() > 71)
{
element = sbuf.substr(70,2);
if (isalpha(element[1]))
{
if (element[0] == ' ')
{
element.erase(0, 1);
elementFound = true;
}
else if (isalpha(element[0]))
{
elementFound = true;
}
}
}
if (!elementFound)
{
stringstream errorMsg;
errorMsg << "WARNING: Problems reading a PDB file\n"
<< " Problems reading a HETATM or ATOM record.\n"
<< " According to the PDB specification,\n"
<< " columns 77-78 should contain the element symbol of an atom.\n"
<< " but OpenBabel found '" << element << "' (atom " << mol.NumAtoms()+1 << ")";
obErrorLog.ThrowError(__FUNCTION__, errorMsg.str(), obWarning);
}
// charge - optional
string scharge;
if (sbuf.size() > 73)
{
scharge = sbuf.substr(72,2);
}
//trim spaces on the right and left sides
while (!atmid.empty() && atmid[0] == ' ')
atmid = atmid.erase(0, 1);
while (!atmid.empty() && atmid[atmid.size()-1] == ' ')
atmid = atmid.substr(0,atmid.size()-1);
/* residue name */
string resname = sbuf.substr(11,3);
if (resname == " ")
resname = "UNK";
else
{
while (!resname.empty() && resname[0] == ' ')
resname = resname.substr(1,resname.size()-1);
while (!resname.empty() && resname[resname.size()-1] == ' ')
resname = resname.substr(0,resname.size()-1);
}
string type;
if (!elementFound) {
// OK, we have to fall back to determining the element from the atom type
// This is unreliable, but there's no other choice
if (EQn(buffer,"ATOM",4)) {
type = atmid.substr(0,2);
if (isdigit(type[0])) {
// sometimes non-standard files have, e.g 11HH
if (!isdigit(type[1])) type = atmid.substr(1,1);
else type = atmid.substr(2,1);
} else if ((sbuf[6] == ' ' &&
strncasecmp(type.c_str(), "Zn", 2) != 0 &&
strncasecmp(type.c_str(), "Fe", 2) != 0) ||
isdigit(type[1])) //type[1] is digit in Platon
type = atmid.substr(0,1); // one-character element
if (resname.substr(0,2) == "AS" || resname[0] == 'N') {
if (atmid == "AD1")
type = "O";
if (atmid == "AD2")
type = "N";
}
if (resname.substr(0,3) == "HIS" || resname[0] == 'H') {
if (atmid == "AD1" || atmid == "AE2")
type = "N";
if (atmid == "AE1" || atmid == "AD2")
type = "C";
}
if (resname.substr(0,2) == "GL" || resname[0] == 'Q') {
if (atmid == "AE1")
type = "O";
if (atmid == "AE2")
type = "N";
}
// fix: #2002557
if (atmid[0] == 'H' &&
(atmid[1] == 'D' || atmid[1] == 'E' ||
atmid[1] == 'G' || atmid[1] == 'H')) // HD, HE, HG, HH, ..
type = "H";
} else { //must be hetatm record
if (isalpha(element[1]) && (isalpha(element[0]) || (element[0] == ' '))) {
if (isalpha(element[0]))
type = element.substr(0,2);
else
type = element.substr(1,1);
if (type.size() == 2)
type[1] = tolower(type[1]);
} else { // no element column to use
if (isalpha(atmid[0])) {
if (atmid.size() > 2 && (atmid[2] == '\0' || atmid[2] == ' '))
type = atmid.substr(0,2);
else if (atmid[0] == 'A') // alpha prefix
type = atmid.substr(1, atmid.size() - 1);
else
type = atmid.substr(0,1);
} else if (atmid[0] == ' ')
type = atmid.substr(1,1); // one char element
else
type = atmid.substr(1,2);
// Some cleanup steps
if (atmid == resname) {
type = atmid;
if (type.size() == 2)
type[1] = tolower(type[1]);
} else
if (resname == "ADR" || resname == "COA" || resname == "FAD" ||
resname == "GPG" || resname == "NAD" || resname == "NAL" ||
resname == "NDP" || resname == "ABA") {
if (type.size() > 1)
type = type.substr(0,1);
//type.erase(1,type.size()-1);
} else // other residues
if (isdigit(type[0])){
type = type.substr(1,1);
}
else
if (type.size() > 1 && isdigit(type[1]))
type = type.substr(0,1);
else
if (type.size() > 1 && isalpha(type[1])) {
if (type[0] == 'O' && type[1] == 'H')
type = type.substr(0,1); // no "Oh" element (e.g. 1MBN)
else if(isupper(type[1])) {
type[1] = tolower(type[1]);
}
}
}
} // HETATM records
} // no element column to use
OBAtom atom;
/* X, Y, Z */
string xstr = sbuf.substr(24,8);
string ystr = sbuf.substr(32,8);
string zstr = sbuf.substr(40,8);
vector3 v(atof(xstr.c_str()),atof(ystr.c_str()),atof(zstr.c_str()));
atom.SetVector(v);
atom.ForceImplH();
// useful for debugging unknown atom types (e.g., PR#1577238)
// cout << mol.NumAtoms() + 1 << " : '" << element << "'" << " " << etab.GetAtomicNum(element.c_str()) << endl;
if (elementFound)
atom.SetAtomicNum(etab.GetAtomicNum(element.c_str()));
else // use our old-style guess from athe atom type
atom.SetAtomicNum(etab.GetAtomicNum(type.c_str()));
if ( (! scharge.empty()) && " " != scharge )
{
if ( isdigit(scharge[0]) && ('+' == scharge[1] || '-' == scharge[1]) )
{
const char reorderCharge[3] = { scharge[1], scharge[0], '\0' };
const int charge = atoi(reorderCharge);
atom.SetFormalCharge(charge);
}
else
{
stringstream errorMsg;
errorMsg << "WARNING: Problems reading a PDB file\n"
<< " Problems reading a HETATM or ATOM record.\n"
<< " According to the PDB specification,\n"
<< " columns 79-80 should contain charge of the atom\n"
<< " but OpenBabel found '" << scharge << "' (atom " << mol.NumAtoms()+1 << ").";
obErrorLog.ThrowError(__FUNCTION__, errorMsg.str(), obWarning);
}
}
else {
atom.SetFormalCharge(0);
}
/* residue sequence number */
string resnum = sbuf.substr(16,4);
OBResidue *res = (mol.NumResidues() > 0) ? mol.GetResidue(mol.NumResidues()-1) : NULL;
if (res == NULL
|| res->GetName() != resname
|| res->GetNumString() != resnum
|| res->GetChain() != chain)
{
vector<OBResidue*>::iterator ri;
for (res = mol.BeginResidue(ri) ; res ; res = mol.NextResidue(ri))
if (res->GetName() == resname
&& res->GetNumString() == resnum
&& static_cast<int>(res->GetChain()) == chain)
break;
if (res == NULL) {
res = mol.NewResidue();
res->SetChain(chain);
res->SetName(resname);
res->SetNum(resnum);
}
}
if (!mol.AddAtom(atom))
return(false);
else {
OBAtom *atom = mol.GetAtom(mol.NumAtoms());
res->AddAtom(atom);
res->SetSerialNum(atom, atoi(serno.c_str()));
res->SetAtomID(atom, sbuf.substr(6,4));
res->SetHetAtom(atom, hetatm);
return(true);
}
} // end reading atom records
bool PDBFormat::WriteMolecule(OBBase* pOb, OBConversion* pConv)
{
OBMol* pmol = dynamic_cast<OBMol*>(pOb);
if(pmol==NULL)
return false;
//Define some references so we can use the old parameter names
ostream &ofs = *pConv->GetOutStream();
OBMol &mol = *pmol;
unsigned int i;
char buffer[BUFF_SIZE];
char type_name[10], padded_name[10];
char the_res[10];
char the_chain = ' ';
const char *element_name;
int res_num;
bool het=true;
int model_num = 0;
if (!pConv->IsLast() || pConv->GetOutputIndex() > 1)
{ // More than one molecule record
model_num = pConv->GetOutputIndex(); // MODEL 1-based index
snprintf(buffer, BUFF_SIZE, "MODEL %8d", model_num);
ofs << buffer << endl;
}
// write back all fields (REMARKS, HELIX, SHEET, SITE, ...)
bool compndWritten = false;
bool authorWritten = false;
std::vector<OBGenericData*> pairData = mol.GetAllData(OBGenericDataType::PairData);
for (std::vector<OBGenericData*>::iterator data = pairData.begin(); data != pairData.end(); ++data) {
OBPairData *pd = static_cast<OBPairData*>(*data);
string attr = pd->GetAttribute();
// filter to make sure we are writing pdb fields only
if (attr != "HEADER" && attr != "OBSLTE" && attr != "TITLE" && attr != "SPLIT" &&
attr != "CAVEAT" && attr != "COMPND" && attr != "SOURCE" && attr != "KEYWDS" &&
attr != "EXPDTA" && attr != "NUMMDL" && attr != "MDLTYP" && attr != "AUTHOR" &&
attr != "REVDAT" && attr != "SPRSDE" && attr != "JRNL" && attr != "REMARK" &&
attr != "DBREF" && attr != "DBREF1" && attr != "DBREF2" && attr != "SEQADV" &&
attr != "SEQRES" && attr != "MODRES" && attr != "HET" && attr != "HETNAM" &&
attr != "HETSYN" && attr != "FORMUL" && attr != "HELIX" && attr != "SHEET" &&
attr != "SSBOND" && attr != "LINK" && attr != "CISPEP" && attr != "SITE" &&
attr != "ORIGX1" && attr != "ORIGX2" && attr != "ORIGX3" && attr != "SCALE1" &&
attr != "SCALE2" && attr != "SCALE3" && attr != "MATRIX1" && attr != "MATRIX2" &&
attr != "MATRIX3" && attr != "MODEL")
continue;
if (attr == "COMPND")
compndWritten = true;
if (attr == "AUTHOR")
authorWritten = true;
// compute spacing needed. HELIX, SITE, HET, ... are trimmed when reading
int nSpacing = 6 - attr.size();
for (int i = 0; i < nSpacing; ++i)
attr += " ";
std::string lines = pd->GetValue();
string::size_type last = 0;
string::size_type pos = lines.find('\n');
while (last != string::npos) {
string line = lines.substr(last, pos - last);
if (pos == string::npos)
last = string::npos;
else
last = pos + 1;
pos = lines.find('\n', last);
ofs << attr << line << endl;
}
}
if (!compndWritten) {
if (strlen(mol.GetTitle()) > 0)
snprintf(buffer, BUFF_SIZE, "COMPND %s ",mol.GetTitle());
else
snprintf(buffer, BUFF_SIZE, "COMPND UNNAMED");
ofs << buffer << endl;
}
if (!authorWritten) {
snprintf(buffer, BUFF_SIZE, "AUTHOR GENERATED BY OPEN BABEL %s",BABEL_VERSION);
ofs << buffer << endl;
}
// Write CRYST1 record, containing unit cell parameters, space group
// and Z value (supposed to be 1)
if (pmol->HasData(OBGenericDataType::UnitCell))
{
OBUnitCell *pUC = (OBUnitCell*)pmol->GetData(OBGenericDataType::UnitCell);
if(pUC->GetSpaceGroup()){
string tmpHM=pUC->GetSpaceGroup()->GetHMName();
// Do we have an extended HM symbol, with origin choice as ":1" or ":2" ? If so, remove it.
size_t n=tmpHM.find(":");
if(n!=string::npos) tmpHM=tmpHM.substr(0,n);
snprintf(buffer, BUFF_SIZE,
"CRYST1%9.3f%9.3f%9.3f%7.2f%7.2f%7.2f %-11s 1",
pUC->GetA(), pUC->GetB(), pUC->GetC(),
pUC->GetAlpha(), pUC->GetBeta(), pUC->GetGamma(),
tmpHM.c_str());
}
else
snprintf(buffer, BUFF_SIZE,
"CRYST1%9.3f%9.3f%9.3f%7.2f%7.2f%7.2f %-11s 1",
pUC->GetA(), pUC->GetB(), pUC->GetC(),
pUC->GetAlpha(), pUC->GetBeta(), pUC->GetGamma(),
"P1");
ofs << buffer << endl;
}
// before we write any records, we should check to see if any coord < -1000
// which will cause errors in the formatting
double minX, minY, minZ;
minX = minY = minZ = -999.0f;
FOR_ATOMS_OF_MOL(a, mol)
{
if (a->GetX() < minX)
minX = a->GetX();
if (a->GetY() < minY)
minY = a->GetY();
if (a->GetZ() < minZ)
minZ = a->GetZ();
}
vector3 transV = VZero;
if (minX < -999.0)
transV.SetX(-1.0*minX - 900.0);
if (minY < -999.0)
transV.SetY(-1.0*minY - 900.0);
if (minZ < -999.0)
transV.SetZ(-1.0*minZ - 900.0);
// if minX, minY, or minZ was never changed, shift will be 0.0f
// otherwise, move enough so that smallest coord is > -999.0f
mol.Translate(transV);
OBAtom *atom;
OBResidue *res;
for (i = 1; i <= mol.NumAtoms(); i++)
{
atom = mol.GetAtom(i);
strncpy(type_name, etab.GetSymbol(atom->GetAtomicNum()), sizeof(type_name));
type_name[sizeof(type_name) - 1] = '\0';
//two char. elements are on position 13 and 14 one char. start at 14
if (strlen(type_name) > 1)
type_name[1] = toupper(type_name[1]);
else
{
char tmp[10];
strncpy(tmp, type_name, 9); // make sure to null-terminate tmp
snprintf(type_name, sizeof(type_name), " %-3s", tmp);
}
if ( (res = atom->GetResidue()) != 0 )
{
het = res->IsHetAtom(atom);
snprintf(the_res,4,"%s",(char*)res->GetName().c_str());
the_res[4] = '\0';
snprintf(type_name,5,"%s",(char*)res->GetAtomID(atom).c_str());
the_chain = res->GetChain();
//two char. elements are on position 13 and 14 one char. start at 14
if (strlen(etab.GetSymbol(atom->GetAtomicNum())) == 1)
{
if (strlen(type_name) < 4)
{
char tmp[10];
strncpy(tmp, type_name, 9); // make sure to null-terminate tmp
snprintf(padded_name, sizeof(padded_name), " %-3s", tmp);
strncpy(type_name,padded_name,4);
type_name[4] = '\0';
}
else
{
/*
type_name[4] = type_name[3];
type_name[3] = type_name[2];
type_name[2] = type_name[1];
type_name[1] = type_name[0];
type_name[0] = type_name[4];
*/
type_name[4] = '\0';
}
}
res_num = res->GetNum();
}
else
{
strcpy(the_res,"UNK");
the_res[3] = '\0';
snprintf(padded_name,sizeof(padded_name), "%s",type_name);
strncpy(type_name,padded_name,4);
type_name[4] = '\0';
res_num = 1;
}
element_name = etab.GetSymbol(atom->GetAtomicNum());
int charge = atom->GetFormalCharge();
char scharge[3] = { ' ', ' ', '\0' };
if(0 != charge)
{
snprintf(scharge, 3, "%+d", charge);
char tmp = scharge[1];
scharge[1] = scharge[0];
scharge[0] = tmp;
}
snprintf(buffer, BUFF_SIZE, "%s%5d %-4s %-3s %c%4d %8.3f%8.3f%8.3f 1.00 0.00 %2s%2s\n",
het?"HETATM":"ATOM ",
i,
type_name,
the_res,
the_chain,
res_num,
atom->GetX(),
atom->GetY(),
atom->GetZ(),
element_name,
scharge);
ofs << buffer;
}
OBAtom *nbr;
vector<OBBond*>::iterator k;
for (i = 1; i <= mol.NumAtoms(); i ++)
{
atom = mol.GetAtom(i);
if (atom->GetValence() == 0)
continue; // no need to write a CONECT record -- no bonds
snprintf(buffer, BUFF_SIZE, "CONECT%5d", i);
ofs << buffer;
// Write out up to 4 real bonds per line PR#1711154
int currentValence = 0;
for (nbr = atom->BeginNbrAtom(k);nbr;nbr = atom->NextNbrAtom(k))
{
snprintf(buffer, BUFF_SIZE, "%5d", nbr->GetIdx());
ofs << buffer;
if (++currentValence % 4 == 0) {
// Add the trailing space to finish this record
ofs << " \n";
// write the start of a new CONECT record
snprintf(buffer, BUFF_SIZE, "CONECT%5d", i);
ofs << buffer;
}
}
// Add trailing spaces
int remainingValence = atom->GetValence() % 4;
for (int count = 0; count < (4 - remainingValence); count++) {
snprintf(buffer, BUFF_SIZE, " ");
ofs << buffer;
}
ofs << " \n";
}
snprintf(buffer, BUFF_SIZE, "MASTER 0 0 0 0 0 0 0 0 ");
ofs << buffer;
snprintf(buffer, BUFF_SIZE, "%4d 0 %4d 0\n",mol.NumAtoms(),mol.NumAtoms());
ofs << buffer;
ofs << "END\n";
if (model_num) {
ofs << "ENDMDL" << endl;
}
return(true);
}
bool PDBFormat::WriteMolecule(OBBase* pOb, OBConversion* pConv)
{
OBMol* pmol = dynamic_cast<OBMol*>(pOb);
if(pmol==NULL)
return false;
//Define some references so we can use the old parameter names
ostream &ofs = *pConv->GetOutStream();
OBMol &mol = *pmol;
unsigned int i;
char buffer[BUFF_SIZE];
char type_name[10], padded_name[10];
char the_res[10];
char the_chain = ' ';
const char *element_name;
int res_num;
bool het=true;
int model_num = 0;
if (!pConv->IsLast() || pConv->GetOutputIndex() > 1)
{ // More than one molecule record
model_num = pConv->GetOutputIndex(); // MODEL 1-based index
snprintf(buffer, BUFF_SIZE, "MODEL %8d", model_num);
ofs << buffer << endl;
}
if (strlen(mol.GetTitle()) > 0)
snprintf(buffer, BUFF_SIZE, "COMPND %s ",mol.GetTitle());
else
snprintf(buffer, BUFF_SIZE, "COMPND UNNAMED");
ofs << buffer << endl;
snprintf(buffer, BUFF_SIZE, "AUTHOR GENERATED BY OPEN BABEL %s",BABEL_VERSION);
ofs << buffer << endl;
// Write CRYST1 record, containing unit cell parameters, space group
// and Z value (supposed to be 1)
if (pmol->HasData(OBGenericDataType::UnitCell))
{
OBUnitCell *pUC = (OBUnitCell*)pmol->GetData(OBGenericDataType::UnitCell);
snprintf(buffer, BUFF_SIZE,
"CRYST1%9.3f%9.3f%9.3f%7.2f%7.2f%7.2f %-11s 1",
pUC->GetA(), pUC->GetB(), pUC->GetC(),
pUC->GetAlpha(), pUC->GetBeta(), pUC->GetGamma(),
pUC->GetSpaceGroup() ?
pUC->GetSpaceGroup()->GetHMName().c_str() : "P1");
ofs << buffer << endl;
}
// before we write any records, we should check to see if any coord < -1000
// which will cause errors in the formatting
double minX, minY, minZ;
minX = minY = minZ = -999.0f;
FOR_ATOMS_OF_MOL(a, mol)
{
if (a->GetX() < minX)
minX = a->GetX();
if (a->GetY() < minY)
minY = a->GetY();
if (a->GetZ() < minZ)
minZ = a->GetZ();
}
vector3 transV = VZero;
if (minX < -999.0)
transV.SetX(-1.0*minX - 900.0);
if (minY < -999.0)
transV.SetY(-1.0*minY - 900.0);
if (minZ < -999.0)
transV.SetZ(-1.0*minZ - 900.0);
// if minX, minY, or minZ was never changed, shift will be 0.0f
// otherwise, move enough so that smallest coord is > -999.0f
mol.Translate(transV);
OBAtom *atom;
OBResidue *res;
for (i = 1; i <= mol.NumAtoms(); i++)
{
atom = mol.GetAtom(i);
strncpy(type_name, etab.GetSymbol(atom->GetAtomicNum()), sizeof(type_name));
type_name[sizeof(type_name) - 1] = '\0';
//two char. elements are on position 13 and 14 one char. start at 14
if (strlen(type_name) > 1)
type_name[1] = toupper(type_name[1]);
else
{
char tmp[10];
strncpy(tmp, type_name, 10);
snprintf(type_name, sizeof(type_name), " %-3s", tmp);
}
if ( (res = atom->GetResidue()) != 0 )
{
het = res->IsHetAtom(atom);
snprintf(the_res,4,"%s",(char*)res->GetName().c_str());
snprintf(type_name,5,"%s",(char*)res->GetAtomID(atom).c_str());
the_chain = res->GetChain();
//two char. elements are on position 13 and 14 one char. start at 14
if (strlen(etab.GetSymbol(atom->GetAtomicNum())) == 1)
{
if (strlen(type_name) < 4)
{
char tmp[16];
strncpy(tmp, type_name, 16);
snprintf(padded_name, sizeof(padded_name), " %-3s", tmp);
strncpy(type_name,padded_name,4);
type_name[4] = '\0';
}
else
{
/*
type_name[4] = type_name[3];
type_name[3] = type_name[2];
type_name[2] = type_name[1];
type_name[1] = type_name[0];
type_name[0] = type_name[4];
*/
type_name[4] = '\0';
}
}
res_num = res->GetNum();
}
else
{
strcpy(the_res,"UNK");
snprintf(padded_name,sizeof(padded_name), "%s",type_name);
strncpy(type_name,padded_name,4);
type_name[4] = '\0';
res_num = 1;
}
element_name = etab.GetSymbol(atom->GetAtomicNum());
int charge = atom->GetFormalCharge();
char scharge[3] = { ' ', ' ', '\0' };
if(0 != charge)
{
snprintf(scharge, 3, "%+d", charge);
char tmp = scharge[1];
scharge[1] = scharge[0];
scharge[0] = tmp;
}
snprintf(buffer, BUFF_SIZE, "%s%5d %-4s %-3s %c%4d %8.3f%8.3f%8.3f 1.00 0.00 %2s%2s\n",
het?"HETATM":"ATOM ",
i,
type_name,
the_res,
the_chain,
res_num,
atom->GetX(),
atom->GetY(),
atom->GetZ(),
element_name,
scharge);
ofs << buffer;
}
OBAtom *nbr;
vector<OBBond*>::iterator k;
for (i = 1; i <= mol.NumAtoms(); i ++)
{
atom = mol.GetAtom(i);
if (atom->GetValence() == 0)
continue; // no need to write a CONECT record -- no bonds
snprintf(buffer, BUFF_SIZE, "CONECT%5d", i);
ofs << buffer;
// Write out up to 4 real bonds per line PR#1711154
unsigned int currentValence = 0;
for (nbr = atom->BeginNbrAtom(k);nbr;nbr = atom->NextNbrAtom(k))
{
unsigned int order = mol.GetBond(atom, nbr)->GetBondOrder();
unsigned int it_order = 0;
for( it_order = 0; it_order < order; it_order++ )
{
if (0 != currentValence && 0 == currentValence % 4)
{
// Add the trailing space to finish this record
ofs << " \n";
// write the start of a new CONECT record
snprintf(buffer, BUFF_SIZE, "CONECT%5d", i);
ofs << buffer;
}
currentValence++;
snprintf(buffer, BUFF_SIZE, "%5d", nbr->GetIdx());
ofs << buffer;
}
}
// Add trailing spaces
unsigned int remainingValence = currentValence % 4;
if( 0 < remainingValence )
{
for (int count = 0; count < (4 - remainingValence); count++)
{
snprintf(buffer, BUFF_SIZE, " ");
ofs << buffer;
}
}
ofs << " \n";
}
snprintf(buffer, BUFF_SIZE, "MASTER 0 0 0 0 0 0 0 0 ");
ofs << buffer;
snprintf(buffer, BUFF_SIZE, "%4d 0 %4d 0\n",mol.NumAtoms(),mol.NumAtoms());
ofs << buffer;
ofs << "END\n";
if (model_num) {
ofs << "ENDMDL" << endl;
}
return(true);
}
