/// <summary>
/// Read a <see cref="IChemFile"/> from a file in PDB format. The molecules
/// in the file are stored as <see cref="IBioPolymer"/>s in the
/// <see cref="IChemFile"/>. The residues are the monomers of the
/// <see cref="IBioPolymer"/>, and their names are the concatenation of the
/// residue, chain id, and the sequence number. Separate chains (denoted by
/// TER records) are stored as separate <see cref="IBioPolymer"/> molecules.
/// </summary>
/// <remarks>
/// Connectivity information is not currently read.
/// </remarks>
/// <returns>The ChemFile that was read from the PDB file.</returns>
private IChemFile ReadChemFile(IChemFile oFile)
{
// initialize all containers
var oSeq = oFile.Builder.NewChemSequence();
var oModel = oFile.Builder.NewChemModel();
var oSet = oFile.Builder.NewAtomContainerSet();
// some variables needed
var oBP = new PDBPolymer();
var molecularStructure = oFile.Builder.NewAtomContainer();
string cRead = "";
char chain = 'A'; // To ensure stringent name giving of monomers
int lineLength = 0;
bool isProteinStructure = false;
atomNumberMap = new Dictionary <int, IAtom>();
if (readConnect.IsSet)
{
bondsFromConnectRecords = new List <IBond>();
}
// do the reading of the Input
try
{
do
{
cRead = oInput.ReadLine();
Debug.WriteLine($"Read line: {cRead}");
if (cRead != null)
{
lineLength = cRead.Length;
// make sure the record name is 6 characters long
if (lineLength < 6)
{
cRead = cRead + " ";
}
// check the first column to decide what to do
var cCol = cRead.Substring(0, 6);
switch (cCol.ToUpperInvariant())
{
case "SEQRES":
{
isProteinStructure = true;
}
break;
case "ATOM ":
{
#region
// read an atom record
var oAtom = ReadAtom(cRead, lineLength);
if (isProteinStructure)
{
// construct a string describing the residue
var cResidue = new StringBuilder(8);
var oObj = oAtom.ResName;
if (oObj != null)
{
cResidue = cResidue.Append(oObj.Trim());
}
oObj = oAtom.ChainID;
if (oObj != null)
{
// cResidue = cResidue.Append(((string)oObj).Trim());
cResidue = cResidue.Append(chain);
}
oObj = oAtom.ResSeq;
if (oObj != null)
{
cResidue = cResidue.Append(oObj.Trim());
}
// search for an existing strand or create a new one.
var strandName = oAtom.ChainID;
if (strandName == null || strandName.Length == 0)
{
strandName = chain.ToString(NumberFormatInfo.InvariantInfo);
}
var oStrand = oBP.GetStrand(strandName);
if (oStrand == null)
{
oStrand = new PDBStrand
{
StrandName = strandName,
Id = chain.ToString(NumberFormatInfo.InvariantInfo)
};
}
// search for an existing monomer or create a new one.
var oMonomer = oBP.GetMonomer(cResidue.ToString(), chain.ToString(NumberFormatInfo.InvariantInfo));
if (oMonomer == null)
{
var monomer = new PDBMonomer
{
MonomerName = cResidue.ToString(),
MonomerType = oAtom.ResName,
ChainID = oAtom.ChainID,
ICode = oAtom.ICode,
ResSeq = oAtom.ResSeq
};
oMonomer = monomer;
}
// add the atom
oBP.AddAtom(oAtom, oMonomer, oStrand);
}
else
{
molecularStructure.Atoms.Add(oAtom);
}
if (readConnect.IsSet)
{
var isDup = atomNumberMap.ContainsKey(oAtom.Serial.Value);
atomNumberMap[oAtom.Serial.Value] = oAtom;
if (isDup)
{
Trace.TraceWarning($"Duplicate serial ID found for atom: {oAtom}");
}
}
Debug.WriteLine($"Added ATOM: {oAtom}");
// As HETATMs cannot be considered to either belong to a certain monomer or strand,
// they are dealt with separately.
#endregion
}
break;
case "HETATM":
{
#region
// read an atom record
var oAtom = ReadAtom(cRead, lineLength);
oAtom.HetAtom = true;
if (isProteinStructure)
{
oBP.Atoms.Add(oAtom);
}
else
{
molecularStructure.Atoms.Add(oAtom);
}
var isDup = atomNumberMap.ContainsKey(oAtom.Serial.Value);
atomNumberMap[oAtom.Serial.Value] = oAtom;
if (isDup)
{
Trace.TraceWarning($"Duplicate serial ID found for atom: {oAtom}");
}
Debug.WriteLine($"Added HETATM: {oAtom}");
#endregion
}
break;
case "TER ":
{
#region
// start new strand
chain++;
var oStrand = new PDBStrand
{
StrandName = chain.ToString(NumberFormatInfo.InvariantInfo)
};
Debug.WriteLine("Added new STRAND");
#endregion
}
break;
case "END ":
{
#region
atomNumberMap.Clear();
if (isProteinStructure)
{
// create bonds and finish the molecule
oSet.Add(oBP);
if (useRebondTool.IsSet)
{
try
{
if (!CreateBondsWithRebondTool(oBP))
{
// Get rid of all potentially created bonds.
Trace.TraceInformation("Bonds could not be created using the RebondTool when PDB file was read.");
oBP.Bonds.Clear();
}
}
catch (Exception exception)
{
Trace.TraceInformation("Bonds could not be created when PDB file was read.");
Debug.WriteLine(exception);
}
}
}
else
{
if (useRebondTool.IsSet)
{
CreateBondsWithRebondTool(molecularStructure);
}
oSet.Add(molecularStructure);
}
#endregion
}
break;
case "MODEL ":
{
#region
// OK, start a new model and save the current one first *if* it contains atoms
if (isProteinStructure)
{
if (oBP.Atoms.Count > 0)
{
// save the model
oSet.Add(oBP);
oModel.MoleculeSet = oSet;
oSeq.Add(oModel);
// setup a new one
oBP = new PDBPolymer();
oModel = oFile.Builder.NewChemModel();
oSet = oFile.Builder.NewAtomContainerSet();
// avoid duplicate atom warnings
atomNumberMap.Clear();
}
}
else
{
if (molecularStructure.Atoms.Count > 0)
{
// save the model
oSet.Add(molecularStructure);
oModel.MoleculeSet = oSet;
oSeq.Add(oModel);
// setup a new one
molecularStructure = oFile.Builder.NewAtomContainer();
oModel = oFile.Builder.NewChemModel();
oSet = oFile.Builder.NewAtomContainerSet();
}
}
#endregion
}
break;
case "REMARK":
{
#region
var comment = oFile.GetProperty <string>(CDKPropertyName.Comment, "");
if (lineLength > 12)
{
comment = comment + cRead.Substring(11).Trim()
+ "\n";
oFile.SetProperty(CDKPropertyName.Comment, comment);
}
else
{
Trace.TraceWarning("REMARK line found without any comment!");
}
#endregion
}
break;
case "COMPND":
{
#region
var title = cRead.Substring(10).Trim();
oFile.SetProperty(CDKPropertyName.Title, title);
#endregion
}
break;
case "CONECT":
{
#region
// Read connectivity information from CONECT records. Only
// covalent bonds are dealt with. Perhaps salt bridges
// should be dealt with in the same way..?
if (!readConnect.IsSet)
{
break;
}
cRead = cRead.Trim();
if (cRead.Length < 16)
{
Debug.WriteLine($"Skipping unexpected empty CONECT line! : {cRead}");
}
else
{
int lineIndex = 6;
int atomFromNumber = -1;
int atomToNumber = -1;
var molecule = (isProteinStructure) ? oBP : molecularStructure;
while (lineIndex + 5 <= cRead.Length)
{
var part = cRead.Substring(lineIndex, 5).Trim();
if (atomFromNumber == -1)
{
try
{
atomFromNumber = int.Parse(part, NumberFormatInfo.InvariantInfo);
}
catch (FormatException)
{
}
}
else
{
try
{
atomToNumber = int.Parse(part, NumberFormatInfo.InvariantInfo);
}
catch (FormatException)
{
atomToNumber = -1;
}
if (atomFromNumber != -1 && atomToNumber != -1)
{
AddBond(molecule, atomFromNumber, atomToNumber);
Debug.WriteLine($"Bonded {atomFromNumber} with {atomToNumber}");
}
}
lineIndex += 5;
}
}
#endregion
}
break;
case "HELIX ":
{
#region
// HELIX 1 H1A CYS A 11 LYS A 18 1 RESIDUE 18 HAS POSITIVE PHI 1D66 72
// 1 2 3 4 5 6 7
// 01234567890123456789012345678901234567890123456789012345678901234567890123456789
var structure = new PDBStructure
{
StructureType = PDBStructure.Helix,
StartChainID = cRead[19],
StartSequenceNumber = int.Parse(cRead.Substring(21, 4).Trim(), NumberFormatInfo.InvariantInfo),
StartInsertionCode = cRead[25],
EndChainID = cRead[31],
EndSequenceNumber = int.Parse(cRead.Substring(33, 4).Trim(), NumberFormatInfo.InvariantInfo),
EndInsertionCode = cRead[37]
};
oBP.Add(structure);
#endregion
}
break;
case "SHEET ":
{
#region
var structure = new PDBStructure
{
StructureType = PDBStructure.Sheet,
StartChainID = cRead[21],
StartSequenceNumber = int.Parse(cRead.Substring(22, 4).Trim(), NumberFormatInfo.InvariantInfo),
StartInsertionCode = cRead[26],
EndChainID = cRead[32],
EndSequenceNumber = int.Parse(cRead.Substring(33, 4).Trim(), NumberFormatInfo.InvariantInfo),
EndInsertionCode = cRead[37]
};
oBP.Add(structure);
#endregion
}
break;
case "TURN ":
{
#region
var structure = new PDBStructure
{
StructureType = PDBStructure.Turn,
StartChainID = cRead[19],
StartSequenceNumber = int.Parse(cRead.Substring(20, 4).Trim(), NumberFormatInfo.InvariantInfo),
StartInsertionCode = cRead[24],
EndChainID = cRead[30],
EndSequenceNumber = int.Parse(cRead.Substring(31, 4).Trim(), NumberFormatInfo.InvariantInfo),
EndInsertionCode = cRead[35]
};
oBP.Add(structure);
#endregion
}
break;
default:
break; // ignore all other commands
}
}
} while (cRead != null);
}
catch (Exception e)
{
if (e is IOException || e is ArgumentException)
{
Trace.TraceError("Found a problem at line:");
Trace.TraceError(cRead);
Trace.TraceError("01234567890123456789012345678901234567890123456789012345678901234567890123456789");
Trace.TraceError(" 1 2 3 4 5 6 7 ");
Trace.TraceError($" error: {e.Message}");
Debug.WriteLine(e);
Console.Error.WriteLine(e.StackTrace);
}
else
{
throw;
}
}
// try to close the Input
try
{
oInput.Close();
}
catch (Exception e)
{
Debug.WriteLine(e);
}
// Set all the dependencies
oModel.MoleculeSet = oSet;
oSeq.Add(oModel);
oFile.Add(oSeq);
return(oFile);
}
