public static IChain ChainFromFileOrCode(string file, PdbLoadOptions loadOptions = PdbLoadOptions.Default, char?chainId = null)
{
IEnumerable <AtomRecord> records = GetPdbRecords(file).Select(record => record as AtomRecord).Where(record => record as HetatmRecord == null);
IChain chain = ChainFromRecords(records, loadOptions, chainId);
return(chain);
}
public static IChain ChainFromText(string text, PdbLoadOptions loadOptions = PdbLoadOptions.Default, char?chainId = null)
{
using (TextReader stream = new StringReader(text))
{
IEnumerable <AtomRecord> records = LoadRecords(stream).Where(record => record is AtomRecord).Cast <AtomRecord>();
IChain chain = ChainFromRecords(records, loadOptions, chainId);
return(chain);
}
}
public static Structure AssemblyFromFileOrCode(string file, PdbLoadOptions loadOptions = PdbLoadOptions.Default)
{
Structure protein = new Structure();
IEnumerable <AtomRecord> records = GetPdbRecords(file, false).Select(r => r as AtomRecord).Where(r => r != null).Where(r => r as HetatmRecord == null);
List <char> letters = records.Select(record => record.ChainId).Distinct().ToList();
foreach (char chainId in records.Select(record => record.ChainId).Distinct())
{
IEnumerable <AtomRecord> chainRecords = records.Where(record => record.ChainId == chainId);
IChain chain = ChainFromRecords(chainRecords);
protein.Add(chain);
}
return(protein);
}
public static Structure AssemblyFromText(string text, PdbLoadOptions loadOptions = PdbLoadOptions.Default)
{
using (TextReader stream = new StringReader(text))
{
IEnumerable <AtomRecord> records = LoadRecords(stream).Where(record => record is AtomRecord).Cast <AtomRecord>();
List <char> letters = records.Select(record => record.ChainId).Distinct().ToList();
Structure protein = new Structure();
foreach (char chainId in letters)
{
IEnumerable <AtomRecord> chainRecords = records.Where(record => record.ChainId == chainId);
IChain chain = ChainFromRecords(chainRecords);
protein.Add(chain);
}
if (protein.Count == 0)
{
return(null);
}
return(protein);
}
}
public static IChain ChainFromRecords(IEnumerable <AtomRecord> records, PdbLoadOptions loadOptions = PdbLoadOptions.Default, char?chainId = null)
{
bool wellFormatted = (loadOptions) == PdbLoadOptions.WellFormatted;
bool requireNCAC = (loadOptions & PdbLoadOptions.RequireNCAC) != 0;
bool requireNonNCAC = (loadOptions & PdbLoadOptions.RequireNonNCAC) != 0;
bool missingAtomsNaN = (loadOptions & PdbLoadOptions.MissingAtomsNaN) != 0;
IEnumerable <AtomSource> allAtoms = records.Where(record => record != null).Select(record => record.ToAtom());
IEnumerable <IGrouping <char, AtomSource> > chains = allAtoms.GroupBy(atom => atom.ChainIndex);
foreach (IGrouping <char, AtomSource> chain in chains)
{
bool chainFailed = false;
char actualChainId = chain.Key;
if (chainId != null && chainId != actualChainId)
{
continue;
}
Chain peptideQuick = new Chain();
List <IGrouping <int, AtomSource> > pdbResidues = chain.GroupBy(atom => atom.ResidueIndex).OrderBy(group => group.Key).ToList();
foreach (IGrouping <int, AtomSource> pdbResidue in pdbResidues)
{
bool nTerminus = pdbResidue == pdbResidues.First();
bool cTerminus = pdbResidue == pdbResidues.Last();
AtomSource N = pdbResidue.FirstOrDefault(atom => atom.Name == "N");
AtomSource CA = pdbResidue.FirstOrDefault(atom => atom.Name == "CA");
AtomSource C = pdbResidue.FirstOrDefault(atom => atom.Name == "C");
// Get the name and create the residue
string name3 = pdbResidue.First().ResidueName;
if (name3 == "HIS") // TODO: Is this the best way of doing this?
{
if (pdbResidue.FirstOrDefault(atom => atom.Name == "HD1" || atom.Name == "1HD") != null)
{
name3 = "HID";
}
else
{
name3 = "HIE";
}
}
// Default to non-disulfide and then replace with disulfides when CYH are nearby with nearby SG atoms
if (name3 == "CYS")
{
name3 = "CYH";
}
if (!AaTable.IsResidueNameKnown(name3))
{
// TODO: Handle ions in some way
if ((new string[] { "ZN", "HOH", "H20", "AU", "CA", "MG", "SO4" }.Contains(name3)))
{
continue;
}
Console.WriteLine("Failing on unknown residue " + name3);
chainFailed = true;
break;
}
IAa residueQuick = new Aa(name3, nTerminus, cTerminus);
// Check for existence of atoms
if (N == null || CA == null || C == null)
{
if (requireNCAC)
{
// TODO: Tracer output
chainFailed = true;
break;
}
}
else if (float.IsNaN(N.XYZ.Length()) || float.IsNaN(CA.XYZ.Length()) || float.IsNaN(C.XYZ.Length()))
{
if (requireNCAC)
{
// TODO: Tracer output
chainFailed = true;
break;
}
}
else
{
residueQuick.AlignToNCAC(N.XYZ, CA.XYZ, C.XYZ);
}
if (missingAtomsNaN)
{
// Only explicitly PDB-defined atoms have real coordinates
for (int i = 0; i < residueQuick.Count; i++)
{
residueQuick[i].Xyz = new Vector3(float.NaN, float.NaN, float.NaN);
}
}
// Method 1
//foreach(AtomQuick atomQuick in residueQuick)
//{
// AtomSource atomSource = pdbResidue.FirstOrDefault(atomPdb => atomPdb.Name == atomQuick.Name);
// // Some PDBs are formatted with hydrogens like 2HD1 instead of HD12
// AtomSource alternateAtomSource = null;
// if(atomQuick.Name.StartsWith("H") && Char.IsNumber(atomQuick.Name.Last())) {
// string alternateName = atomQuick.Name.Last() + atomQuick.Name.Substring(0, atomQuick.Name.Length - 1);
// alternateAtomSource = pdbResidue.FirstOrDefault(atomPdb => atomPdb.Name == alternateName);
// }
// if(atomSource != null)
// {
// atomQuick.XYZ = atomSource.XYZ;
// }
// else if (alternateAtomSource != null)
// {
// atomQuick.XYZ = alternateAtomSource.XYZ;
// }
// else if (!allowMissingAtoms)
// {
// // TODO ... return null?
// continue;
// }
//}
// Method 2 - explicit atom name alternatives
// First pass: initialize atom locations based on exact name-matched PDB records
List <IAtom> initializedAtoms = new List <IAtom>();
List <AtomSource> usedSourceAtoms = new List <AtomSource>();
List <IAtom> uninitializedAtoms = new List <IAtom>(residueQuick);
foreach (IAtom atomQuick in residueQuick)
{
AtomSource atomSource = pdbResidue.FirstOrDefault(atomPdb => atomPdb.Name == atomQuick.Name);
if (atomSource != null)
{
atomQuick.Xyz = atomSource.XYZ;
initializedAtoms.Add(atomQuick);
uninitializedAtoms.Remove(atomQuick);
usedSourceAtoms.Add(atomSource);
}
}
// Second pass: try to initialize any remaining atoms from PDB records corresponding to known alternative atom names. Multiple
// passes are made to try to resolve the names, moving from more to less preferred.
int alternateAtomNamesMaxCount = alternateAtomNames_.Values.Select(value => value.Count).Max();
for (int nameIndex = 0; nameIndex < alternateAtomNamesMaxCount; nameIndex++)
{
foreach (Atom atomQuick in residueQuick)
{
if (initializedAtoms.Contains(atomQuick))
{
continue;
}
List <string> alternativeAtomNames = null;
if (!alternateAtomNames_.TryGetValue(atomQuick.Name, out alternativeAtomNames))
{
continue;
}
if (alternativeAtomNames.Count <= nameIndex)
{
continue;
}
IEnumerable <AtomSource> atomSources = pdbResidue.Where(atomPdb => alternativeAtomNames[nameIndex] == atomPdb.Name);
foreach (AtomSource atomSource in atomSources)
{
if (usedSourceAtoms.Contains(atomSource))
{
continue;
}
atomQuick.Xyz = atomSource.XYZ;
initializedAtoms.Add(atomQuick);
uninitializedAtoms.Remove(atomQuick);
usedSourceAtoms.Add(atomSource);
}
}
}
if (initializedAtoms.Count != residueQuick.Count)
{
if (!requireNCAC && !requireNonNCAC)
{
continue;
}
IEnumerable <IAtom> uninitializedNCAC = uninitializedAtoms.Where(atom => atom.Name == "N" || atom.Name == "CA" || atom.Name == "C").Where(atom => float.IsNaN(atom.Xyz.Length()));
IEnumerable <IAtom> uninitializedNonNCAC = uninitializedAtoms.Where(atom => atom.Name != "N" && atom.Name != "CA" && atom.Name != "C").Where(atom => float.IsNaN(atom.Xyz.Length()));
if (requireNCAC && uninitializedNCAC.Count() > 0)
{
chainFailed = true;
break;
}
if (requireNonNCAC && uninitializedNonNCAC.Count() > 0)
{
chainFailed = true;
break;
}
//// TODO: Create trace output and move this there
//Console.WriteLine("Failed to initialize residue " + residueQuick.Name);
//foreach (AtomQuick atom in uninitializedAtoms)
//{
// Console.WriteLine(atom.Name);
//}
//chainFailed = true;
//break;
}
peptideQuick.Add(residueQuick);
}
if (!chainFailed && peptideQuick.Count > 0)
{
AddDisulfides(peptideQuick);
return(peptideQuick);
}
}
return(null);
}
