public virtual void InitMeasure(List <string> fileNames, string alignFile, bool referenceFlag, string juryProfile = null)
{
this.alignFile = alignFile;
if (referenceFlag)
{
jury = new jury1D();
jury.PrepareJury(fileNames, alignFile, juryProfile);
//AddErrors(jury.errors);
}
}
public virtual void InitMeasure(string profilesFile, bool referenceFlag, string juryProfile = null, string refJuryProfile = null)
{
this.alignFile = profilesFile;
if (referenceFlag)
{
jury = new jury1D();
jury.PrepareJury(profilesFile, refJuryProfile);
//AddErrors(jury.errors);
}
}
public virtual void InitMeasure(string dirName, string alignFile, bool referenceFlag, string juryProfile = null)
{
this.alignFile = alignFile;
this.dirName = dirName;
if (referenceFlag)
{
jury = new jury1D();
jury.PrepareJury(dirName, alignFile, juryProfile);
//AddErrors(jury.errors);
}
}
public virtual void InitMeasure(DCDFile dcd, string alignFile, bool referenceFlag, string juryProfile = null)
{
this.alignFile = alignFile;
if (referenceFlag)
{
if (juryProfile.Length == 0 || !File.Exists(juryProfile))
{
throw new Exception("Profile for reference structure searching has been not defined or does not exist");
}
jury = new jury1D();
jury.PrepareJury(dcd, alignFile, juryProfile);
}
}
private StatClust Calc(ClusterOutput outp, bool flag)
{
List <string> five = new List <string>();
List <string> listFiles = new List <string>();
//if (outp == null || outp.clusters == null)
// return null;
if (outp == null || outp.dirName == null)
{
return(null);
}
string [] aux = outp.dirName.Split(Path.DirectorySeparatorChar);
string native = aux[aux.Length - 1] + ".pdb";
string dirM = "";
for (int i = 0; i < aux.Length - 1; i++)
{
dirM += aux[i] + Path.DirectorySeparatorChar;
}
native = dirM + native;
if (!File.Exists(native))
{
return(null);
}
five.Add(native);
jury1D jury;
if (outp.clusters != null)
{
outp.clusters.Sort((a, b) => b.Count.CompareTo(a.Count));
int end = outp.clusters.Count;
if (outp.clusters.Count > 5)
{
end = 5;
}
for (int i = 0; i < end; i++)
{
listFiles.Clear();
foreach (var item in outp.clusters[i])
{
listFiles.Add(outp.dirName + Path.DirectorySeparatorChar + item);
}
//
//
//Sift sift = new Sift(listFiles);
//ClusterOutput oo = sift.Shape();
if (flag)
{
DistanceMeasure distRmsd = new Rmsd(listFiles, "", false, uQlustCore.PDB.PDBMODE.ONLY_CA);
distRmsd.InitMeasure();
string strName = distRmsd.GetReferenceStructure(outp.clusters[i]);
five.Add(outp.dirName + Path.DirectorySeparatorChar + strName);
}
else
{
jury = new jury1D();
//jury.PrepareJury(listFiles, "", "Z:\\dcd\\SS3_SA9_jury_internal.profiles");
ClusterOutput oo = jury.JuryOptWeights(outp.clusters[i]);
if (oo == null)
{
continue;
}
five.Add(outp.dirName + Path.DirectorySeparatorChar + oo.juryLike[0].Key);
}
//five.Add(outp.dirName + Path.DirectorySeparatorChar + outp.clusters[i][0]);
}
}
if (outp.hNode != null)
{
int end;
List <List <string> > cli = outp.hNode.GetClusters(10);
cli.Sort((a, b) => b.Count.CompareTo(a.Count));
outp.clusters = new List <List <string> >();
end = cli.Count;
if (cli.Count > 5)
{
end = 5;
}
for (int i = 0; i < end; i++)
{
// listFiles.Clear();
//foreach (var item in cli[i])
//{
//listFiles.Add(
//}
listFiles.Clear();
foreach (var item in cli[i])
{
listFiles.Add(outp.dirName + Path.DirectorySeparatorChar + item);
}
jury = new jury1D();
jury.PrepareJury(listFiles, "", "C:\\data\\dcd\\SS8_SA3_jury_internal.profiles");
outp.clusters.Add(cli[0]); //jury
ClusterOutput oo = jury.JuryOptWeights(cli[i]);
if (oo == null)
{
continue;
}
five.Add(outp.dirName + Path.DirectorySeparatorChar + cli[i][0]);
//five.Add(outp.dirName + Path.DirectorySeparatorChar + strName);
}
}
if (outp.juryLike != null)
{
int end = outp.juryLike.Count;
if (outp.juryLike.Count > 5)
{
end = 5;
}
for (int i = 0; i < end; i++)
{
five.Add(outp.dirName + Path.DirectorySeparatorChar + outp.juryLike[i].Key);
}
}
// DistanceMeasure dist;
// dist = new Rmsd(five, "", false, PDB.PDBMODE.ONLY_CA);
// dist = new MaxSub(five, "", false);
Dictionary <string, double> cc = ReadScore(aux[aux.Length - 1]);
if (cc == null)
{
return(null);
}
StatClust stCLust = new StatClust();
string[] tt1 = native.Split(Path.DirectorySeparatorChar);
stCLust.native = tt1[tt1.Length - 1];
for (int i = 1; i < five.Count; i++)
{
string[] tt2 = five[i].Split(Path.DirectorySeparatorChar);
// double rmsd = dist.GetDistance(tt1[tt1.Length-1], tt2[tt2.Length-1]) /100.0;
double rmsd = 0;
if (cc.ContainsKey(tt2[tt2.Length - 1]))
{
rmsd = cc[tt2[tt2.Length - 1]];
}
if (rmsd > stCLust.rmsd)
{
if (outp.juryLike != null)
{
stCLust.size = outp.juryLike.Count;
}
else
{
stCLust.size = outp.clusters[i - 1].Count;
}
stCLust.rmsd = rmsd;
string[] dd = five[i].Split(Path.DirectorySeparatorChar);
stCLust.reference = dd[dd.Length - 1];
}
}
if (stCLust.rmsd < 1000)
{
avrSum += stCLust.rmsd;
avrCount++;
}
if (outp.hNode != null)
{
outp.clusters.Clear();
outp.clusters = null;
}
return(stCLust);
}
private void Run1DJury(string name, string dirName, string alignFile=null, DCDFile dcd = null)
{
DateTime cpuPart1 = DateTime.Now;
ClusterOutput output;
jury1D ju=new jury1D();
progressDic.Add(name,ju);
//DistanceMeasure distance = CreateMeasure();
if (opt.other.alignGenerate)
opt.other.alignFileName = "";
if (alignFile != null)
ju.PrepareJury(alignFile, opt.other.juryProfile);
else
if (dcd != null)
ju.PrepareJury(dcd, alignFile, opt.other.juryProfile);
else
ju.PrepareJury(dirName, alignFile, opt.other.juryProfile);
clType = ju.ToString();
DateTime cpuPart2 = DateTime.Now;
//jury1D ju = new jury1D(opt.weightHE,opt.weightC,(JuryDistance) distance);
//output = ju.JuryOpt(new List<string>(ju.stateAlign.Keys));
if (ju.alignKeys != null)
{
output = ju.JuryOptWeights(ju.alignKeys);
}
else
{
UpadateJobInfo(name, true, false);
throw new Exception("Alignment is epmty! Check errors");
}
UpdateOutput(name, dirName,alignFile, output,ju.ToString(), cpuPart1,cpuPart2, ju);
}
public ClusterOutput Run(List<string> structs)
{
maxProgress=5;
currentProgress=0;
if (hConcensus)
{
maxProgress++;
consensus = new HammingConsensus(dMeasure.dirName, null, false, input.consensusProfile);
progressObject = consensus;
consensus.InitMeasure();
currentProgress += 1.0 / maxProgress;
}
jury = new jury1D();
progressObject = jury;
currentProgress += 1.0 / maxProgress;
progressObject = null;
jury.PrepareJury(dMeasure.dirName, dMeasure.alignFile, input.jury1DProfileFast);
currentProgress += 1.0 / maxProgress;
ClusterOutput clOut = new ClusterOutput();
root.setStruct = structs;
// if(hConcensus)
// consensus.ToConsensusStates(structs);
FastCluster(root);
maxV = initNodesNum;
while (st.Count>0 && (leaves.Count+st.Count)<initNodesNum)
{
st.Sort(
delegate(HClusterNode p1, HClusterNode p2)
{
return p2.setStruct.Count.CompareTo(p1.setStruct.Count);
}
);
HClusterNode node = st[0];
st.RemoveAt(0);
FastCluster(node);
currentV += leaves.Count + st.Count;
}
currentV = maxV;
currentProgress+=1.0/maxProgress;
while (st.Count > 0)
{
HClusterNode node = st[0];
st.RemoveAt(0);
leaves.Add(node);
}
MakeDendrogs(linkage);
currentProgress+=1.0/maxProgress;
PrepareList();
root = ConnectDendrogs(linkage);
root.levelDist = root.SearchMaxDist();
root.realDist = dMeasure.GetRealValue(root.levelDist);
//CheckRefDistances();
//dendrogList = RearangeDendrogram(root);
//root = ConnectDendrogs();
clOut.hNode = root;
currentProgress+=1.0/maxProgress;
return clOut;
}
public void CalcStat()
{
List<string> fileNames = new List<string>();
List<string> refStructures = new List<string>();
DistanceMeasures measure = distanceControl1.distDef;
DistanceMeasure distTemp = null;
double distRes=0;
maxV = selected.Count;
distList.Clear();
try
{
tableRes = new DataTable();
tableRes.Columns.Add("Cluster Size", typeof(int));
tableRes.Columns.Add("Reference structure", typeof(string));
tableRes.Columns.Add("Distance", typeof(double));
tableRes.Columns.Add("Hidden1", typeof(int));
tableRes.Columns.Add("Hidden2", typeof(string));
if (checkBoxSable.Checked)
{
tableRes.Columns.Add("Dist to Sable", typeof(double));
maxV *= 2;
}
if (checkBox1.Checked)
{
for (int i = 0; i < selected.Count; i++)
maxV += selected[i].Count;
}
fileNames.Add(textBox1.Text);
jury1D jury = new jury1D();
if (distanceControl1.reference)
jury.PrepareJury(dirName, alignFile, distanceControl1.referenceProfile);
refStructures.Clear();
selected.Sort(delegate(List<string> first, List<string> second)
{ return first.Count.CompareTo(second.Count); });
selected.Reverse();
// dataGridView1.Rows.Add(selected.Count);
for (int i = 0; i < selected.Count; i++)
{
string refD = selected[i][0];
ClusterOutput juryO = null;
if (distanceControl1.reference)
{
if (selected[i].Count > 5)
{
juryO = jury.JuryOptWeights(selected[i]);
if (juryO == null)
continue;
refD = juryO.juryLike[0].Key;
if (!fileNames.Contains(dirName + Path.DirectorySeparatorChar + juryO.juryLike[0].Key))
{
fileNames.Add(dirName + Path.DirectorySeparatorChar + juryO.juryLike[0].Key);
refStructures.Add(dirName + Path.DirectorySeparatorChar + refD);
}
}
else
if (!fileNames.Contains(dirName + Path.DirectorySeparatorChar + selected[i][0]))
{
fileNames.Add(dirName + Path.DirectorySeparatorChar + selected[i][0]);
refStructures.Add(dirName + Path.DirectorySeparatorChar + selected[i][0]);
}
}
else
{
dist = PrepareDistanceMeasure(selected[i], measure, dirName);
distTemp = dist;
refD = dist.GetReferenceStructure(selected[i]);
fileNames.Add(dirName + Path.DirectorySeparatorChar + refD);
refStructures.Add(dirName + Path.DirectorySeparatorChar + refD);
}
if (fileNames.Count == 2)
{
switch (measure)
{
case DistanceMeasures.HAMMING:
dist = new JuryDistance(fileNames, alignFile, false, distanceControl1.profileName);
break;
case DistanceMeasures.MAXSUB:
dist = new MaxSub(fileNames, null, false);
break;
case DistanceMeasures.RMSD:
dist = new Rmsd(fileNames, null, false, distanceControl1.CAtoms);
break;
}
dist.InitMeasure();
distRes = Convert.ToDouble(String.Format("{0:0.00}", dist.GetDistance(native, refD) / 100.0));
}
fileNames.RemoveAt(fileNames.Count - 1);
if (checkBoxSable.Checked && refStructures.Count > 0)
tableRes.Rows.Add(selected[i].Count, refD, distRes, i, dirName, 0.0);
else
tableRes.Rows.Add(selected[i].Count, refD, distRes, i, dirName);
currentV++;
}
if (checkBoxSable.Checked && refStructures.Count > 0)
{
Dictionary<string, double> res = null;// SableDist(refStructures);
for (int i = 0; i < refStructures.Count; i++)
{
if (res.ContainsKey(Path.GetFileName(refStructures[i])))
{
DataRow dr = tableDist.Rows[i];
dr[5] = Convert.ToDouble(String.Format("{0:0.00}", res[Path.GetFileName(refStructures[i])]));
}
}
}
if (checkBox1.Checked)
{
CalculateDistToAll();
}
}
catch(Exception ex)
{
exc = ex;
}
}
protected Dictionary <byte,int>[] JuryConsensusStates(Alignment alUse,List<string> structNames)
{
jury1D juryLocal = new jury1D();
juryLocal.PrepareJury(alUse,stateAlign);
Dictionary <byte ,int>[] cons=new Dictionary<byte,int>[columns.Length];
ClusterOutput output = juryLocal.JuryOptWeights(structNames);
if (output == null)
return null;
int i = 0;
string result;
if (structNames.Contains("consensus"))
result = "consensus";
else
result = output.juryLike[0].Key;
foreach (var item in stateAlign[result])
{
cons[i] = new Dictionary<byte, int>();
cons[i++].Add(item, 0);
}
return cons;
}
public ClusterOutput AutomaticCluster(List<string> _structNames,bool dendrog=false)
{
HashCInput remInput = input;
int remRelClusters=0;
int remPerData = 0;
ClusterOutput output = new ClusterOutput();
ClusterOutput[] backOutput = new ClusterOutput[10];
Dictionary<int, Dictionary<int, double>> res = new Dictionary<int, Dictionary<int, double>>();
double finalRes = Double.MinValue;
PrepareClustering(_structNames);
dicC = SelectColumnsByEntropy(dicC, dicC.Keys.Count - dicC.Keys.Count/5, 0.95);
List<string> keys =new List<string>(dicC.Keys);
Dictionary<string,string> backDic=new Dictionary<string,string>();
foreach (var item in _structNames)
allItems.Add(dirName + Path.DirectorySeparatorChar + item);
for (int i = 0; i < backOutput.Length; i++)
backOutput[i] = new ClusterOutput();
jury = new jury1D();
jury.PrepareJury(allItems, null, input.profileName);
Dictionary<string, List<int>> remDic = new Dictionary<string, List<int>>(dicC);
for(int i=2;i<10;i++)//relevant clusters
{
if (output.clusters != null)
{
output.clusters.Clear();
foreach(var item in backOutput)
item.clusters.Clear();
}
input.relClusters = i;
Dictionary<int, double> aux = new Dictionary<int, double>();
for (int j = 30; j < 90; j += 5) //percentage
{
dicC = remDic;
input.perData = j;
output.clusters = PrepareClusters(dicC, structNames);
double backDisp = 0;
List<double> backList = new List<double>();
foreach (var item in backOutput)
{
List<string> kk = new List<string>(remDic.Keys);
KeyValuePair<Dictionary<string, List<int>>, List<string>> backGround = default(KeyValuePair<Dictionary<string, List<int>>, List<string>>);
backGround = MakeBackgroundData(kk);
backDic.Clear();
foreach (var bitem in backGround.Key)
backDic.Add(backGround.Value[bitem.Value[0]], bitem.Key);
dicC = backGround.Key;
item.clusters = PrepareClusters(backGround.Key, backGround.Value);
backList.Add(Math.Log(CalcDisp(item.clusters, backDic)));
backDisp += backList[backList.Count-1];
}
backDisp /= backOutput.Length;
double disp = 0;
disp = Math.Log(CalcDisp(output.clusters, structToKey));
double sdk = 0;
foreach(var item in backList)
{
sdk += (item - backDisp) * (item - backDisp);
}
sdk = Math.Sqrt(1.0 / backList.Count * sdk)*Math.Sqrt(1.0/backList.Count+1);
double gap = backDisp - disp;
aux.Add(j, gap);
// aux.Add(j, CalculateDaviesBouldinIndex(output.clusters));
if (j>30 && aux[j-5]-aux[j]+sdk >= finalRes)
{
finalRes = aux[j - 5] - aux[j] + sdk;
remRelClusters = i;
remPerData = j-5;
//break;
}
}
res.Add(i, aux);
}
input.relClusters = remRelClusters;
input.perData = remPerData;
output.clusters = PrepareClusters(dicC, structNames);
input = remInput;
return output;
}
List<KeyValuePair<string,double>> CLusterRepresentJury(string dirName, List<string> targets,string profileName)
{
List<string> fileNames = new List<string>(targets.Count);
foreach (var item in targets)
fileNames.Add(dirName + Path.DirectorySeparatorChar + item);
jury1D jury = new jury1D();
jury.PrepareJury(fileNames, null, profileName);
ClusterOutput opt = jury.JuryOptWeights(targets);
return opt.juryLike;
}
public void CalcStat()
{
List <string> fileNames = new List <string>();
List <string> refStructures = new List <string>();
DistanceMeasures measure = distanceControl1.distDef;
DistanceMeasure distTemp = null;
double distRes = 0;
maxV = selected.Count;
distList.Clear();
try
{
tableRes = new DataTable();
tableRes.Columns.Add("Cluster Size", typeof(int));
tableRes.Columns.Add("Reference structure", typeof(string));
tableRes.Columns.Add("Distance", typeof(double));
tableRes.Columns.Add("Hidden1", typeof(int));
tableRes.Columns.Add("Hidden2", typeof(string));
if (checkBoxSable.Checked)
{
tableRes.Columns.Add("Dist to Sable", typeof(double));
maxV *= 2;
}
if (checkBox1.Checked)
{
for (int i = 0; i < selected.Count; i++)
{
maxV += selected[i].Count;
}
}
fileNames.Add(textBox1.Text);
jury1D jury = new jury1D();
if (distanceControl1.reference)
{
jury.PrepareJury(dirName, alignFile, distanceControl1.referenceProfile);
}
refStructures.Clear();
selected.Sort(delegate(List <string> first, List <string> second)
{ return(first.Count.CompareTo(second.Count)); });
selected.Reverse();
// dataGridView1.Rows.Add(selected.Count);
for (int i = 0; i < selected.Count; i++)
{
string refD = selected[i][0];
ClusterOutput juryO = null;
if (distanceControl1.reference)
{
if (selected[i].Count > 5)
{
juryO = jury.JuryOptWeights(selected[i]);
if (juryO == null)
{
continue;
}
refD = juryO.juryLike[0].Key;
if (!fileNames.Contains(dirName + Path.DirectorySeparatorChar + juryO.juryLike[0].Key))
{
fileNames.Add(dirName + Path.DirectorySeparatorChar + juryO.juryLike[0].Key);
refStructures.Add(dirName + Path.DirectorySeparatorChar + refD);
}
}
else
if (!fileNames.Contains(dirName + Path.DirectorySeparatorChar + selected[i][0]))
{
fileNames.Add(dirName + Path.DirectorySeparatorChar + selected[i][0]);
refStructures.Add(dirName + Path.DirectorySeparatorChar + selected[i][0]);
}
}
else
{
dist = PrepareDistanceMeasure(selected[i], measure, dirName);
distTemp = dist;
refD = dist.GetReferenceStructure(selected[i]);
fileNames.Add(dirName + Path.DirectorySeparatorChar + refD);
refStructures.Add(dirName + Path.DirectorySeparatorChar + refD);
}
if (fileNames.Count == 2)
{
switch (measure)
{
case DistanceMeasures.HAMMING:
dist = new JuryDistance(fileNames, alignFile, false, distanceControl1.profileName);
break;
case DistanceMeasures.MAXSUB:
dist = new MaxSub(fileNames, null, false);
break;
case DistanceMeasures.RMSD:
dist = new Rmsd(fileNames, null, false, distanceControl1.CAtoms);
break;
}
dist.InitMeasure();
distRes = Convert.ToDouble(String.Format("{0:0.00}", dist.GetDistance(native, refD) / 100.0));
}
fileNames.RemoveAt(fileNames.Count - 1);
if (checkBoxSable.Checked && refStructures.Count > 0)
{
tableRes.Rows.Add(selected[i].Count, refD, distRes, i, dirName, 0.0);
}
else
{
tableRes.Rows.Add(selected[i].Count, refD, distRes, i, dirName);
}
currentV++;
}
if (checkBoxSable.Checked && refStructures.Count > 0)
{
Dictionary <string, double> res = null;// SableDist(refStructures);
for (int i = 0; i < refStructures.Count; i++)
{
if (res.ContainsKey(Path.GetFileName(refStructures[i])))
{
DataRow dr = tableDist.Rows[i];
dr[5] = Convert.ToDouble(String.Format("{0:0.00}", res[Path.GetFileName(refStructures[i])]));
}
}
}
if (checkBox1.Checked)
{
CalculateDistToAll();
}
}
catch (Exception ex)
{
exc = ex;
}
}
public ClusterOutput DendrogUsingMeasures(List<string> structures)
{
jury1D juryLocal = new jury1D();
juryLocal.PrepareJury(al);
ClusterOutput outC = null;
Dictionary<string, List<int>> dic;
//Console.WriteLine("Start after jury " + Process.GetCurrentProcess().PeakWorkingSet64);
maxV = 4;
currentV = 0;
dic = PrepareKeys(structures,false);
currentV++;
//DebugClass.DebugOn();
//dic = HashEntropyCombine(dic, structures,input.reqClusters);
//Console.WriteLine("Entropy ready after jury " + Process.GetCurrentProcess().PeakWorkingSet64);
DebugClass.WriteMessage("Entropy ready");
//Alternative way to start of UQclust Tree must be finished
//input.relClusters = 10000;
input.relClusters = input.reqClusters;
input.perData = 90;
//dic = FastCombineKeys(dic, structures, false);
dic = FastCombineKeys(dic, structures, true);
DebugClass.WriteMessage("dic size" + dic.Count);
currentV++;
//Console.WriteLine("Combine ready after jury " + Process.GetCurrentProcess().PeakWorkingSet64);
DebugClass.WriteMessage("Combine Keys ready");
Dictionary<string, string> translateToCluster = new Dictionary<string, string>(dic.Count);
List<string> structuresToDendrogram = new List<string>(dic.Count);
List<string> structuresFullPath = new List<string>(dic.Count);
DebugClass.WriteMessage("Number of clusters: "+dic.Count);
int cc = 0;
foreach (var item in dic)
{
if (item.Value.Count > 2)
{
List<string> cluster = new List<string>(item.Value.Count);
foreach (var str in item.Value)
cluster.Add(structures[str]);
ClusterOutput output = juryLocal.JuryOptWeights(cluster);
structuresToDendrogram.Add(output.juryLike[0].Key);
if(alignFile==null)
structuresFullPath.Add(dirName + Path.DirectorySeparatorChar + output.juryLike[0].Key);
else
structuresFullPath.Add(output.juryLike[0].Key);
translateToCluster.Add(output.juryLike[0].Key, item.Key);
}
else
{
structuresToDendrogram.Add(structures[item.Value[0]]);
if(alignFile==null)
structuresFullPath.Add(dirName + Path.DirectorySeparatorChar + structures[item.Value[0]]);
else
structuresFullPath.Add(structures[item.Value[0]]);
translateToCluster.Add(structures[item.Value[0]], item.Key);
}
cc++;
}
currentV++;
DebugClass.WriteMessage("Jury finished");
switch (dMeasure)
{
case DistanceMeasures.HAMMING:
if (refJuryProfile == null || !jury1d)
throw new Exception("Sorry but for jury measure you have to define 1djury profile to find reference structure");
else
dist = new JuryDistance(structuresFullPath, alignFile, true, profileName, refJuryProfile);
break;
case DistanceMeasures.COSINE:
dist = new CosineDistance(structuresFullPath, alignFile, jury1d, profileName, refJuryProfile);
break;
case DistanceMeasures.RMSD:
dist = new Rmsd(structuresFullPath, "", jury1d, atoms, refJuryProfile);
break;
case DistanceMeasures.MAXSUB:
dist = new MaxSub(structuresFullPath, "", jury1d, refJuryProfile);
break;
}
// return new ClusterOutput();
DebugClass.WriteMessage("Start hierarchical");
//Console.WriteLine("Start hierarchical " + Process.GetCurrentProcess().PeakWorkingSet64);
currentV = maxV;
hk = new hierarchicalCluster(dist, linkageType, dirName);
dist.InitMeasure();
//Now just add strctures to the leaves
outC = hk.HierarchicalClustering(structuresToDendrogram);
DebugClass.WriteMessage("Stop hierarchical");
List<HClusterNode> hLeaves = outC.hNode.GetLeaves();
foreach(var item in hLeaves)
{
if (translateToCluster.ContainsKey(item.setStruct[0]))
{
foreach (var str in dic[translateToCluster[item.setStruct[0]]])
if (item.setStruct[0] != structures[str])
item.setStruct.Add(structures[str]);
}
else
throw new Exception("Cannot add structure. Something is wrong");
}
outC.hNode.RedoSetStructures();
return outC;
}
private void button2_Click(object sender, EventArgs e)
{
string[] aux;
if(checkBox1.Checked)
{
if(jury1DSetup1.profileName==null || jury1DSetup1.profileName.Length==0)
{
MessageBox.Show("Profile name for 1djury must be specified!");
this.DialogResult = DialogResult.None;
return;
}
jury1D jury = new jury1D();
jury.PrepareJury(structures, "",jury1DSetup1.profileName);
List <string> prep=new List<string>();
foreach (var item in structures)
{
aux = item.Split(Path.DirectorySeparatorChar);
prep.Add(aux[aux.Length - 1]);
}
ClusterOutput oc=jury.JuryOptWeights(prep);
for(int i=0;i<selectBest1.bestNumber;i++)
bestJuryStructures.Add(oc.juryLike[i].Key);
}
this.DialogResult = DialogResult.OK;
if (selectBest1.getFileName != null && File.Exists(selectBest1.getFileName))
{
DistanceMeasures measure = selectBest1.measure;
structures.Add(selectBest1.getFileName);
DistanceMeasure dist = null;
switch (measure)
{
case DistanceMeasures.HAMMING:
dist = new JuryDistance(structures, "", false, selectBest1.hammingProfile);
break;
case DistanceMeasures.MAXSUB:
dist = new MaxSub(structures, "", false);
break;
case DistanceMeasures.RMSD:
dist = new Rmsd(structures, "", false, selectBest1.CAtoms);
break;
}
List<KeyValuePair<string, int>> distList = new List<KeyValuePair<string, int>>();
aux = selectBest1.getFileName.Split(Path.DirectorySeparatorChar);
string native = aux[aux.Length - 1];
foreach (var item in structures)
{
aux = item.Split(Path.DirectorySeparatorChar);
int val = dist.GetDistance(native, aux[aux.Length - 1]);
distList.Add(new KeyValuePair<string, int>(aux[aux.Length - 1], val));
}
distList.Sort((firstPair, nextPair) =>
{
return firstPair.Value.CompareTo(nextPair.Value);
});
for (int i = 0; i < selectBest1.bestNumber; i++)
{
bestStructures.Add(distList[i].Key);
}
}
}
private ClusterOutput PrepareClustersJuryLike(Dictionary<string, List<int>> dic, List<string> structures)
{
List<string> juryList = new List<string>();
ClusterOutput juryOut;
jury1D jury = new jury1D();
jury.PrepareJury(al);
foreach (var item in dic.Keys)
{
List<string> aux = new List<string>(dic[item].Count);
foreach (var st in dic[item])
aux.Add(structures[st]);
if (aux.Count > 3)
{
juryOut = jury.JuryOptWeights(aux);
juryList.Add(dirName+Path.DirectorySeparatorChar+juryOut.juryLike[0].Key);
}
else
juryList.Add(dirName + Path.DirectorySeparatorChar + aux[0]);
}
Sift sift = new Sift();
juryOut=sift.RunSift(juryList);
//juryOut = jury.JuryOptWeights(juryList);
return juryOut;
}
private StatClust Calc(ClusterOutput outp,bool flag)
{
List <string> five=new List<string>();
List<string> listFiles = new List<string>();
//if (outp == null || outp.clusters == null)
// return null;
if(outp==null || outp.dirName==null)
return null;
string []aux=outp.dirName.Split(Path.DirectorySeparatorChar);
string native=aux[aux.Length-1]+".pdb";
string dirM = "";
for (int i = 0; i < aux.Length - 1; i++)
dirM += aux[i] + Path.DirectorySeparatorChar;
native = dirM + native;
if(!File.Exists(native))
return null;
five.Add(native);
// jury1D jury = new jury1D(outp.dirName, "", "C:\\data\\dcd\\SS_SA9internal.profiles");
// jury1D jury = new jury1D(outp.dirName, "", "C:\\data\\dcd\\SS3_SA9_jury_internal.profiles");
//jury1D jury = new jury1D(outp.dirName, "", "C:\\data\\dcd\\SS8_SA3_jury_internal.profiles");
jury1D jury;
// jury1D jury = new jury1D(outp.dirName, "", "C:\\data\\dcd\\SS_Contact_jury_internal.profiles");
if(outp.clusters!=null)
{
outp.clusters.Sort((a,b)=>b.Count.CompareTo(a.Count));
int end = outp.clusters.Count;
if (outp.clusters.Count > 5)
end = 5;
for (int i = 0; i < end; i++)
{
listFiles.Clear();
foreach (var item in outp.clusters[i])
{
listFiles.Add(outp.dirName + Path.DirectorySeparatorChar+item);
}
//
//
//Sift sift = new Sift(listFiles);
//ClusterOutput oo = sift.Shape();
if (flag)
{
DistanceMeasure distRmsd = new Rmsd(listFiles, "", false, uQlustCore.PDB.PDBMODE.ONLY_CA);
distRmsd.InitMeasure();
string strName = distRmsd.GetReferenceStructure(outp.clusters[i]);
five.Add(outp.dirName + Path.DirectorySeparatorChar + strName);
}
else
{
jury = new jury1D();
jury.PrepareJury(listFiles, "", "Z:\\dcd\\SS3_SA9_jury_internal.profiles");
ClusterOutput oo = jury.JuryOptWeights(outp.clusters[i]);
if (oo == null)
continue;
five.Add(outp.dirName + Path.DirectorySeparatorChar + oo.juryLike[0].Key);
}
//five.Add(outp.dirName + Path.DirectorySeparatorChar + outp.clusters[i][0]);
}
}
if (outp.hNode != null)
{
int end;
List<List<string>> cli=outp.hNode.GetClusters(10);
cli.Sort((a, b) => b.Count.CompareTo(a.Count));
outp.clusters = new List<List<string>>();
end = cli.Count;
if (cli.Count > 5)
end = 5;
for (int i = 0; i < end; i++)
{
// listFiles.Clear();
//foreach (var item in cli[i])
//{
//listFiles.Add(
//}
listFiles.Clear();
foreach (var item in cli[i])
{
listFiles.Add(outp.dirName + Path.DirectorySeparatorChar + item);
}
jury = new jury1D();
jury.PrepareJury(listFiles, "", "C:\\data\\dcd\\SS8_SA3_jury_internal.profiles");
outp.clusters.Add(cli[0]); //jury
ClusterOutput oo = jury.JuryOptWeights(cli[i]);
if (oo == null)
continue;
five.Add(outp.dirName + Path.DirectorySeparatorChar + cli[i][0]);
//five.Add(outp.dirName + Path.DirectorySeparatorChar + strName);
}
}
if (outp.juryLike != null)
{
int end = outp.juryLike.Count;
if (outp.juryLike.Count > 5)
end = 5;
for (int i = 0; i < end; i++)
five.Add(outp.dirName + Path.DirectorySeparatorChar + outp.juryLike[i].Key);
}
// DistanceMeasure dist;
// dist = new Rmsd(five, "", false, PDB.PDBMODE.ONLY_CA);
// dist = new MaxSub(five, "", false);
Dictionary<string, double> cc = ReadScore(aux[aux.Length - 1]);
if (cc == null)
return null ;
StatClust stCLust = new StatClust();
string[] tt1 = native.Split(Path.DirectorySeparatorChar);
stCLust.native = tt1[tt1.Length - 1];
for (int i = 1; i < five.Count; i++)
{
string[] tt2 = five[i].Split(Path.DirectorySeparatorChar);
// double rmsd = dist.GetDistance(tt1[tt1.Length-1], tt2[tt2.Length-1]) /100.0;
double rmsd=0;
if (cc.ContainsKey(tt2[tt2.Length-1]))
rmsd = cc[tt2[tt2.Length - 1]];
if (rmsd > stCLust.rmsd)
{
if (outp.juryLike != null)
stCLust.size = outp.juryLike.Count;
else
stCLust.size = outp.clusters[i - 1].Count;
stCLust.rmsd = rmsd;
string[] dd = five[i].Split(Path.DirectorySeparatorChar);
stCLust.reference = dd[dd.Length-1];
}
}
if (stCLust.rmsd < 1000)
{
avrSum += stCLust.rmsd;
avrCount++;
}
if (outp.hNode != null)
{
outp.clusters.Clear();
outp.clusters = null;
}
return stCLust;
}
private void button2_Click(object sender, EventArgs e)
{
string[] aux;
if (checkBox1.Checked)
{
if (jury1DSetup1.profileName == null || jury1DSetup1.profileName.Length == 0)
{
MessageBox.Show("Profile name for 1djury must be specified!");
this.DialogResult = DialogResult.None;
return;
}
jury1D jury = new jury1D();
jury.PrepareJury(structures, "", jury1DSetup1.profileName);
List <string> prep = new List <string>();
foreach (var item in structures)
{
aux = item.Split(Path.DirectorySeparatorChar);
prep.Add(aux[aux.Length - 1]);
}
ClusterOutput oc = jury.JuryOptWeights(prep);
for (int i = 0; i < selectBest1.bestNumber; i++)
{
bestJuryStructures.Add(oc.juryLike[i].Key);
}
}
this.DialogResult = DialogResult.OK;
if (selectBest1.getFileName != null && File.Exists(selectBest1.getFileName))
{
DistanceMeasures measure = selectBest1.measure;
structures.Add(selectBest1.getFileName);
DistanceMeasure dist = null;
switch (measure)
{
case DistanceMeasures.HAMMING:
dist = new JuryDistance(structures, "", false, selectBest1.hammingProfile);
break;
case DistanceMeasures.MAXSUB:
dist = new MaxSub(structures, "", false);
break;
case DistanceMeasures.RMSD:
dist = new Rmsd(structures, "", false, selectBest1.CAtoms);
break;
}
List <KeyValuePair <string, int> > distList = new List <KeyValuePair <string, int> >();
aux = selectBest1.getFileName.Split(Path.DirectorySeparatorChar);
dist.InitMeasure();
string native = aux[aux.Length - 1];
foreach (var item in structures)
{
aux = item.Split(Path.DirectorySeparatorChar);
int val = dist.GetDistance(native, aux[aux.Length - 1]);
distList.Add(new KeyValuePair <string, int>(aux[aux.Length - 1], val));
}
distList.Sort((firstPair, nextPair) =>
{
return(firstPair.Value.CompareTo(nextPair.Value));
});
for (int i = 0; i < selectBest1.bestNumber; i++)
{
bestStructures.Add(distList[i].Key);
}
}
}
