/// <summary>
/// This method parses a PDB formatted file for a molecular primary structure.
/// It will only parse the first model in the file (as defined by the ENDMDL record. Additional models in the PDB file will be discarded.
/// </summary>
/// <param name="filename"></param>
/// <returns></returns>
public static PrimaryStructure GetPrimaryStructure(string filename)
{
PrimaryStructure model;
StreamReader sr = null;
try {
sr = new StreamReader(filename);
model = new PrimaryStructure();
bool EOF = false;
// all indexing is 1 based and will increment before first assignment
int atomIndex = 0;
int residueIndex = 0;
int chainIndex = 0;
Chain chain = null;
Residue residue = null;
int currentResidueID = -1;
string currentChainID = "";
bool TERFound = false;
while (!EOF)
{
string record = sr.ReadLine();
// Console.WriteLine("processing Line: " + record);
if (record == null || record.Trim().Equals("END"))
{
EOF = true;
}
else
{
if (record == null || record.Trim() == "")
{
continue;
}
string recordName = record;
if (recordName.Length < 6)
{
recordName = recordName.Trim();
}
else
{
recordName = record.Substring(0, 6).Trim();
}
switch (recordName)
{
case "ATOM":
case "HETATM":
int atomResidueID = int.Parse(record.Substring(22, 4).Trim());
if (atomResidueID != currentResidueID)
{
// store residue - first time here will be null
if (residue != null)
{
model.AddResidue(residueIndex, residue);
if (chain != null && currentChainID != "")
{
chain.AddResidue(residueIndex, residue);
}
}
residueIndex++;
currentResidueID = atomResidueID;
string residueName = record.Substring(17, 3).Trim();
residue = new Residue(residueIndex, atomResidueID, residueName);
}
string chainID = record.Substring(21, 1).Trim();
if (currentChainID != chainID || TERFound)
{
Console.WriteLine("Handling new chain");
if (chain != null && currentChainID != "")
{
model.AddChain(chain);
}
if (chainID != null)
{
chainIndex++;
chain = new Chain(chainIndex, chainID);
}
currentChainID = chainID;
TERFound = false;
}
int atomID = int.Parse(record.Substring(6, 5).Trim());
string atomName = record.Substring(12, 4).Trim();
Element element = ElementHelper.Parse(record.Substring(76, 2).Trim());
float charge = 0;
// if the file has extra columns at the end then assume they are for charge and try to parse
if (record.Length > 80)
{
string chargeString = record.Substring(80, record.Length - 80).Trim();
if (chargeString.Length > 0)
{
try {
charge = float.Parse(chargeString);
}
catch (Exception) {
// do nothing
}
}
}
// PDB coordinates are in angstroms. Convert to nanometres.
Vector3 position = new Vector3();
position.x = float.Parse(record.Substring(30, 8)) / 10;
position.y = float.Parse(record.Substring(38, 8)) / 10;
position.z = float.Parse(record.Substring(46, 8)) / 10;
atomIndex++;
Atom atom = new Atom(atomIndex, atomID, atomName, element, position, charge);
//Debug.Log("Atom: " + atom);
// check for and store main chain elements.
if (recordName.Equals("ATOM") && residue.ResidueType != StandardResidue.None)
{
switch (atom.Name)
{
case "N":
residue.AmineNitrogen = atom;
break;
case "CA":
residue.AlphaCarbon = atom;
break;
case "C":
residue.CarbonylCarbon = atom;
break;
case "O":
case "O1":
case "OT1":
residue.CarbonylOxygen = atom;
break;
}
}
atom.ResidueIndex = residueIndex;
atom.ResidueID = currentResidueID;
atom.ResidueName = residue.Name;
atom.ResidueType = residue.ResidueType;
if (currentChainID != "")
{
atom.ChainID = currentChainID;
}
residue.Atoms.Add(atomIndex, atom);
model.AddAtom(atomIndex, atom);
break;
case "TER":
Console.WriteLine("Found TER");
TERFound = true;
break;
case "TITLE":
if (model.Title == null)
{
model.Title = "";
}
// many PDB's have long remark sections. Abbreviate for now.
if (model.Title.Length < 200)
{
model.Title += record.Substring(7) + "\n";
}
break;
case "ENDMDL":
case "END":
// currently only parse the first model in the file
EOF = true;
break;
}
}
}
if (residue != null)
{
model.AddResidue(residueIndex, residue);
if (chain != null)
{
chain.AddResidue(residueIndex, residue);
}
}
if (chain != null && currentChainID != "")
{
model.AddChain(chain);
}
}
catch (Exception e) {
throw new FileParseException(e.Message);
}
finally {
if (sr != null)
{
sr.Close();
}
}
return(model);
}
private static PrimaryStructure GetFrame(StreamReader sr)
{
PrimaryStructure model;
// todo: improve file format error handling for misformatted files
try {
model = new PrimaryStructure();
string titleLine = sr.ReadLine();
// remove any blank lines before title line
while (titleLine != null && titleLine.Trim().Equals(""))
{
titleLine = sr.ReadLine();
}
if (titleLine == null)
{
return(null);
}
model.Title = titleLine;
model.Time = 0;
Regex g = new Regex(@"\st=\s*(\d+\.?\d*)");
Match m = g.Match(titleLine);
if (m.Success)
{
string frameTimeString = m.Groups[1].Value;
model.Time = float.Parse(frameTimeString);
// Console.WriteLine("Success parsing[" + titleLine + "] for frame time");
}
else
{
// no frame time in header. Assume structure file, leave frame time at 0;
// return false;
}
int residueIndex = 0;
int chainIndex = 0;
Chain chain = null;
Residue residue = null;
int currentResidueID = -1;
Residue lastResidue = null;
int atomCount = Int32.Parse(sr.ReadLine().Trim());
for (int i = 0; i < atomCount; i++)
{
string atomLine = sr.ReadLine();
if (atomLine == null) // if something is wrong with the file, i.e. ends before total atom count
{
break;
}
// if atom residue number not the existing residue number store the existing residue and instantiate new residue
int atomResidueID = int.Parse(atomLine.Substring(0, 5).Trim());
if (atomResidueID != currentResidueID)
{
// store residue - first time here will be null
if (residue != null)
{
model.AddResidue(residueIndex, residue);
if (chain != null)
{
chain.AddResidue(residueIndex, residue);
}
}
residueIndex++;
currentResidueID = atomResidueID;
string residueName = atomLine.Substring(5, 5).Trim();
lastResidue = residue;
residue = new Residue(residueIndex, currentResidueID, residueName);
// Residue type changes used to capture chain information. May not be 100% accurate but not parsing topology files so options are limited.
// Also checking to see if last two atoms before end of amino acid were Oxygen, signifying a chain terminator
if ((lastResidue == null || residue.ResidueType != lastResidue.ResidueType) ||
(lastResidue.ResidueType == StandardResidue.AminoAcid && model.Atoms()[i - 1].Element == Element.O && model.Atoms()[i - 2].Element == Element.O))
{
if (chain != null)
{
model.AddChain(chain);
}
chainIndex++;
chain = new Chain(chainIndex, chainIndex.ToString());
}
}
int atomIndex = i;
int atomID = int.Parse((atomLine.Substring(15, 5)).Trim());
string atomName = atomLine.Substring(10, 5).Trim();
Element element = ElementHelper.Parse(atomName);
Vector3 position = new Vector3();
position.x = float.Parse(atomLine.Substring(20, 8));
position.y = float.Parse(atomLine.Substring(28, 8));
position.z = float.Parse(atomLine.Substring(36, 8));
Atom atom = new Atom(atomIndex, atomID, atomName, element, position);
// check for and store main chain elements.
if (residue.ResidueType != StandardResidue.None)
{
switch (atom.Name)
{
case "N":
residue.AmineNitrogen = atom;
break;
case "CA":
residue.AlphaCarbon = atom;
break;
case "C":
residue.CarbonylCarbon = atom;
break;
case "O":
case "O1":
case "OT1":
residue.CarbonylOxygen = atom;
break;
}
}
atom.ResidueIndex = residueIndex;
atom.ResidueID = currentResidueID;
atom.ResidueName = residue.Name;
atom.ResidueType = residue.ResidueType;
atom.ChainID = chainIndex.ToString();
residue.Atoms.Add(atomIndex, atom);
model.AddAtom(atomIndex, atom);
}
if (residue != null)
{
model.AddResidue(residueIndex, residue);
if (chain != null)
{
chain.AddResidue(residueIndex, residue);
}
}
if (chain != null)
{
model.AddChain(chain);
}
// Parse box vectors.
// todo: test and make this more robust for various string lengths
float[] vertices = parseBoxLine(sr.ReadLine());
model.OriginalBoundingBox = new BoundingBox(vertices[0], vertices[1], vertices[2], vertices[3], vertices[4], vertices[5], vertices[6], vertices[7], vertices[8]);
}
catch (Exception e) {
throw new FileParseException(e.Message);
}
return(model);
}