/// <summary>
/// Read input data from database
/// </summary>
/// <param name="smp">
/// <returns></returns>
List<CompoundStructureActivityData> ReadData(
SasMapParms smp)
{
MetaColumn activityMc = smp.EndpointMc;
QueryColumn keyCriteriaQc = smp.KeyCriteriaQc;
AssertMx.IsNotNull(activityMc, "mc");
AssertMx.IsNotNull(keyCriteriaQc, "keyCriteriaQc");
MetaTable mt, mt2;
MetaColumn mc2 = null;
Query q = new Query();
mt = activityMc.MetaTable;
QueryTable qt = new QueryTable(mt);
if (mt.SummarizedExists && !mt.UseSummarizedData)
{ // retrieve summarized data if exists
mt2 = MetaTableCollection.Get(mt.Name + MetaTable.SummarySuffix);
if (mt2 != null)
{
mc2 = mt2.GetMetaColumnByName(activityMc.Name);
if (mc2 == null) mc2 = mt2.GetMetaColumnByLabel(activityMc.Label);
}
if (mc2 != null) // same column available in summarized?
{
mt = mt2;
activityMc = mc2;
}
}
SMP.KeyCriteriaQc.CopyCriteriaToQueryKeyCritera(q);
q.KeyCriteriaDisplay = SMP.KeyCriteriaQc.CriteriaDisplay;
qt.SelectKeyOnly();
QueryColumn qc = qt.GetQueryColumnByName(activityMc.Name);
qc.Selected = true;
q.AddQueryTable(qt);
QueryEngine qe = new QueryEngine();
List<string> keyList = qe.ExecuteQuery(q); // note that keylist may be empty if single-step query
HashSet<string> keySet = new HashSet<string>(StringComparer.OrdinalIgnoreCase);
List<CompoundStructureActivityData> data = new List<CompoundStructureActivityData>();
int rowCount = 0;
while (true)
{
object[] vo = qe.NextRow();
if (vo == null) break;
CompoundStructureActivityData cd = new CompoundStructureActivityData();
string cid = (string)vo[0];
cd.Cid = cid;
keySet.Add(cid); // accumulate keys
object val = vo[2];
if (NullValue.IsNull(val)) continue;
if (val is double)
cd.Activity = (double)val;
else if (val is Int32)
cd.Activity = (Int32)val;
else if (val is NumberMx)
{
NumberMx nex = val as NumberMx;
cd.Activity = nex.Value;
}
else if (val is QualifiedNumber)
{
QualifiedNumber qn = val as QualifiedNumber;
cd.Activity = qn.NumberValue;
//if (qn.Qualifier != null && qn.Qualifier != "" && qn.Qualifier != "=")
// continue; // (don't want to do this since may filter out good data (e.g. IC50 <0.0001))
}
else continue;
if (cd.Activity == NullValue.NullNumber) continue;
data.Add(cd);
rowCount++;
}
// Retrieve structures
keyList = new List<string>(keySet);
Dictionary<string, MoleculeMx> csDict = MoleculeUtil.SelectMoleculesForCidList(keyList, qt.MetaTable); // get the structures in a single step
// Add structures and build/store fingerprints to data
DebugLog.Message("========== Fingerprints ============");
foreach (CompoundStructureActivityData cd in data)
{
if (!csDict.ContainsKey(cd.Cid) || csDict[cd.Cid] == null) continue;
if (cd.Cid == "111" || cd.Cid == "222") csDict = csDict; // debug
MoleculeMx cs = csDict[cd.Cid];
cd.Structure = cs;
FingerprintType fpType = FingerprintType.Circular;
int fpSubtype = -1;
if (SMP.SimilarityType == SimilaritySearchType.ECFP4) // some issue with ECFP4?
{
fpType = FingerprintType.Circular;
fpSubtype = CircularFingerprintType.ECFP4;
}
else if (SMP.SimilarityType == SimilaritySearchType.Normal)
{
fpType = FingerprintType.MACCS;
}
cd.BitsetFingerprint = cs.BuildBitSetFingerprint(fpType, fpSubtype);
if (cd.BitsetFingerprint == null) continue; // couldn't build fingerprint (e.g. no structure)
if (Debug) DebugLog.Message(cd.Cid + ": " + Lex.Join(CdkMolUtil.GetBitSet(cd.BitsetFingerprint), ", "));
}
return data;
}
/// <summary>
/// NextRow - Return the next matching row value object
/// </summary>
/// <returns></returns>
public override Object[] NextRow()
{
throw new NotImplementedException();
#if false
string cid = "";
Molecule substituent = null;
Stopwatch swTotal = Stopwatch.StartNew();
Stopwatch sw = Stopwatch.StartNew();
if (Eqp.SearchKeySubset == null || Eqp.SearchKeySubset.Count == 0)
{
return(null);
}
// Get the structures for the set of keys
if (CidToStructureDict == null)
{
CidToStructureDict = MoleculeUtil.SelectMoleculesForCidList(Eqp.SearchKeySubset);
KeyListPos = -1;
MapPos = -1;
int msTime = (int)sw.ElapsedMilliseconds;
if (RGroupDecomp.Debug)
{
DebugLog.Message("Select " + Eqp.SearchKeySubset.Count + " structures time(ms): " + msTime);
}
}
// Get the next match
while (true)
{
if (KeyListPos < 0 || MapPos + 1 >= MapCount)
{ // go to next structure & set up mapping
KeyListPos++;
if (KeyListPos >= Eqp.SearchKeySubset.Count)
{
return(null);
}
cid = Eqp.SearchKeySubset[KeyListPos];
if (!CidToStructureDict.ContainsKey(cid))
{
continue;
}
MoleculeMx cs = CidToStructureDict[cid];
string molFile = cs.GetMolfileString();
//string chime = cs.ChimeString; // debug
if (RGroupDecomp.Debug)
{
DebugLog.StopwatchMessage("t1", sw);
}
if (DebugMx.False) // debug
{
//molFile = FileUtil.ReadFile(@"C:\Downloads\RGroupTarget.mol");
string targetSmiles = "CC1Cn2cc(c(= O)c3c2c(c(c(c3)F)Cl)S1)C(= O)O";
string targetChime = "CYAAFQwADfwQ19aXPcZERR45lQkn08$hZNXzeJ2yaAhDnxxJou4Gq9od8VG1ykiO63fQpvM8W4C6MR$O3VaZjQwrGr5weW3y^BeUEezndoIivvAbQN58EEHVMAsdPaF4LIsqsf$OCBUPHI5njBB2LIBy3i2cwbrD8T8kFVBVWkTCfIUFUtblI0G7vYiEL^svUWCT^m6tF18I7ISJUp^7WkuzpT9LrBSJLmMl5hHXog$68Q6YPb0^Xp0ftxmy7FDSF^sWib6^JrUMhrHtLfJ3yVMTm9RIrvvKMRQvKqq4G1Ooze5pdlpSdzp7MFl0K1zx4tdnNFoUO1kRPpyZks61qbBz2tU0L$svvojoU4yUlf$^MFF0nqSfGVW2PKv9TReb$knLXffdAAhqBN310WUfdJDoQBNX1a5L2uj9ybNNRLYpaZN1p6WYp2WI^ntQEVTBaJF1Uu28N4o2xudURpVITTKO7omUtpgLoaoOQHazmZG3k^aHPUOQfE0d27eAbE^uxcQUAB";
molFile = CdkMol.Util.MoleculeTofMolfileString(CdkMol.Util.ChimeStringToMolecule(targetChime));
}
Molecule target = CdkMol.Util.MolfileStringToMolecule(molFile);
if (RGroupDecomp.Debug)
{
DebugLog.StopwatchMessage("t2", sw);
}
MapCount = RGroupDecomp.ProcessTargetMolecule(target); // process & get number of maps
if (MapCount > 1)
{
MapCount = MapCount; // debug, seems to always be 1
}
MapPos = -1;
if (RGroupDecomp.Debug)
{
DebugLog.StopwatchMessage("t3", sw);
}
}
// Get the substituent for each RGroup for the next mapping
MapPos++; // get data for next mapping
for (int si = 0; si < Substituents.Length; si++) // clear substituents
{
Substituents[si] = null;
}
for (int fi = 0; fi < RgTotalCount; fi++)
{
Molecule fragment = RGroupDecomp.GetIthMappingFragment(MapPos, fi);
if (fragment == null)
{
break; // must have reached the end
}
if (RGroupDecomp.Debug)
{
DebugLog.StopwatchMessage("t4", sw);
}
int ri = CdkMol.Util.GetFragmentRGroupAssignment(fragment);
if (ri >= 0 && ri < Substituents.Length)
{
Substituents[ri] = fragment;
}
}
if (RGroupDecomp.Debug)
{
DebugLog.StopwatchMessage("t5", sw);
}
// Fill in the Vo
object[] vo = new object[Qt.SelectedCount];
for (int ci = 0; ci < SelectList.Count; ci++)
{
MetaColumn mc = SelectList[ci];
string name = mc.Name.ToLower();
if (mc.IsKey)
{
vo[ci] = Eqp.SearchKeySubset[KeyListPos];
}
else if (Lex.Eq(name, "Core"))
{
vo[ci] = CoreChemicalStructure;
}
else if (Lex.Eq(name, "Terminate_Option"))
{
vo[ci] = TerminateOptionString;
}
else if (Lex.Eq(name, "Map_Number"))
{
vo[ci] = (MapPos + 1).ToString();
}
else if (Lex.Eq(name, "Map"))
{
if (RGroupDecomp.Debug)
{
DebugLog.StopwatchMessage("t6", sw);
}
MoleculeMx cs = null;
try // hilight core molecule
{
if (StrMatcher == null) // initialize matcher with core structure if not done yet
{
StrMatcher = new StructureMatcher();
string molfile2 = CdkMol.Util.RemoveRGroupAttachmentPointAtoms(CoreChemicalStructure.GetMolfileString());
if (RGroupDecomp.Debug)
{
DebugLog.StopwatchMessage("t7", sw);
}
MoleculeMx cs2 = new MoleculeMx(MoleculeFormat.Molfile, molfile2);
StrMatcher.SetSSSQueryMolecule(cs2); // set core query used for highlighting
if (RGroupDecomp.Debug)
{
DebugLog.StopwatchMessage("t8", sw);
}
}
Molecule alignedTarget = RGroupDecomp.GetAlignedTargetForMapping(MapPos);
if (RGroupDecomp.Debug)
{
DebugLog.StopwatchMessage("t9", sw);
}
string chime = CdkMol.Util.MoleculeToChimeString(alignedTarget);
if (RGroupDecomp.Debug)
{
DebugLog.StopwatchMessage("t10", sw);
}
cs = new MoleculeMx(MoleculeFormat.Chime, chime);
cs = StrMatcher.HighlightMatchingSubstructure(cs);
}
catch (Exception ex) { ex = ex; }
vo[ci] = cs; // store value
if (RGroupDecomp.Debug)
{
DebugLog.StopwatchMessage("t11", sw);
}
}
else if (Lex.Eq(name, "Is_Map_Complete"))
{
vo[ci] = "Yes"; // always complete for PP
//else vo[ci] = "No";
}
else
{ // must be a substituent
if (RGroupDecomp.Debug)
{
DebugLog.StopwatchMessage("t12", sw);
}
int ri = name.IndexOf("_");
if (ri < 0 || !int.TryParse(name.Substring(1, ri - 1), out ri))
{
continue; // in case of bogus name
}
substituent = Substituents[ri];
if (substituent == null)
{
vo[ci] = null; // no mapping found
continue;
}
if (name.EndsWith("_structure"))
{
string chime = CdkMol.Util.MoleculeToChimeString(substituent);
MoleculeMx cs = new MoleculeMx(MoleculeFormat.Chime, chime);
vo[ci] = cs;
}
else if (name.EndsWith("_smiles"))
{
string smiles = CdkMol.Util.MoleculeToSmilesString(substituent);
vo[ci] = smiles;
}
else if (name.EndsWith("_formula"))
{
string mf = CdkMol.Util.GetMolFormulaDotDisconnect(substituent);
vo[ci] = mf;
}
else if (name.EndsWith("_weight"))
{
vo[ci] = CdkMol.Util.GetMolWeight(substituent);
}
else if (name.EndsWith("_substno"))
{
vo[ci] = ri;
}
if (RGroupDecomp.Debug)
{
DebugLog.StopwatchMessage("t13", sw);
}
}
}
int msTime = (int)sw.ElapsedMilliseconds;
if (RGroupDecomp.Debug)
{
DebugLog.StopwatchMessage("Total Time for Cid: " + cid, swTotal);
}
return(vo);
}
#endif
}