/// <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>
/// Search button clicked, process input
/// </summary>
/// <returns></returns>
DialogResult ProcessInput()
{
CidList cidList = null;
StreamReader structureFileReader = null;
string qid; // query identifier, compoundId, file name or sdFile key value
QueryManager qm;
DataTableMx dt;
DataColumn dc;
DataRowMx dr;
//object[] dr; // if using Qe
DialogResult dlgRslt = DialogResult.OK;
Query q = null;
QueryTable qt = null;
QueryColumn simScoreQc, structQc; // query column containing latest query settings
MetaTable mt = null;
MetaColumn keyMc = null, structMc = null, dbSetMc = null, simScoreMc = null, mc;
MetaColumnType storageType;
string txt, tok;
if (DatabasesToSearch.Text == "")
{
MessageBoxMx.ShowError("Databases to search must be defined.");
DatabasesToSearch.Focus();
return(DialogResult.Cancel);
}
// Get list of databases
string[] dba = DatabasesToSearch.Text.Split(',');
List <MetaTable> tables = new List <MetaTable>();
foreach (string dbLabel0 in dba)
{
string dbLabel = dbLabel0.Trim();
RootTable dbInfo = RootTable.GetFromTableLabel(dbLabel);
if (dbInfo == null)
{
MessageBoxMx.ShowError("Can't find database " + DatabasesToSearch.Text);
DatabasesToSearch.Focus();
return(DialogResult.Cancel);
}
mt = MetaTableCollection.Get(dbInfo.MetaTableName);
if (mt == null)
{
MessageBoxMx.ShowError("Unable to locate parent structure table for database: " + DatabasesToSearch.Text);
DatabasesToSearch.Focus();
return(DialogResult.Cancel);
}
if (dbSetMc == null)
{
dbSetMc = mt.DatabaseListMetaColumn;
}
tables.Add(mt);
}
if (dbSetMc == null)
{
throw new Exception("\"Databases\" metacolumn not found for any of the databases to search");
}
// Validate other form values
RetrieveStructures = RetrieveMatchingStructures.Checked;
bool fromList = FromList.Checked;
int listCidsRead = 0;
int inputQueryCount = -1;
if (fromList) // using list, validate list name
{
if (SavedListUo == null)
{
MessageBoxMx.ShowError("Compound list must be defined.");
ListName.Focus();
return(DialogResult.Cancel);
}
cidList = CidListCommand.Read(SavedListUo);
if (cidList == null)
{
MessageBoxMx.ShowError("Error reading list.");
ListName.Focus();
return(DialogResult.Cancel);
}
inputQueryCount = cidList.Count;
}
else // Using SdFile, validate SdFile name
{
StructureFile = FileName.Text;
if (StructureFile == "")
{
MessageBoxMx.ShowError("File must be defined.");
FileName.Focus();
return(DialogResult.Cancel);
}
try { structureFileReader = new StreamReader(StructureFile); structureFileReader.Close(); }
catch (Exception ex)
{
MessageBoxMx.ShowError("Can't read file: " + Lex.Dq(StructureFile));
FileName.Focus();
return(DialogResult.Cancel);
}
keyField = KeyField.Text; // get key, blank to use name in 1st line
inputQueryCount = -1; // don't know how many queries unless we read the file (todo?)
}
tok = ResultsName.Text.Trim(); // name to store results under
if (tok == "")
{
MessageBoxMx.ShowError("A name for the results must be provided.");
ResultsName.Focus();
return(DialogResult.Cancel);
}
if (SubStruct.Checked)
{
Psc.SearchType = StructureSearchType.Substructure;
}
else if (Full.Checked)
{
Psc.SearchType = StructureSearchType.FullStructure;
}
else if (Similarity.Checked)
{
Psc.SearchType = StructureSearchType.MolSim;
}
else
{
throw new Exception("Unrecognized search type");
}
// Write initial log entries
SearchCount++;
string logFileName = ClientDirs.DefaultMobiusUserDocumentsFolder + @"\Multistructure Search " + SearchCount + ".txt";
if (!UIMisc.CanWriteFileToDefaultDir(logFileName))
{
return(DialogResult.Cancel);
}
LogStream = new StreamWriter(logFileName);
if (ResultsUo == null)
{
ResultsUo = new UserObject(UserObjectType.Annotation);
}
ResultsUo.Name = tok;
UserObjectTree.GetValidUserObjectTypeFolder(ResultsUo);
DateTime startTime = DateTime.Now;
WriteToLog("Multiple " + Psc.SearchType + " Search");
WriteToLog("Databases: " + DatabasesToSearch.Text);
WriteToLog("Date: " + startTime);
if (fromList)
{
WriteToLog("Input List: " + SavedListUo.Name);
}
else
{
WriteToLog("Input Structure File: " + StructureFile);
}
WriteToLog("Output List: " + ResultsUo.Name);
WriteToLog("Log File: " + logFileName);
WriteToLog("");
WriteToLog("Query, Match, Score");
int queryCount = 0;
int matchAtLeastOneCount = 0;
MoleculeMx queryStructure = null; // current structure being searched
CidList matchList = new CidList();
List <MatchData> matchData = new List <MatchData>();
if (FromFile.Checked) // open SdFile as required
{
structureFileReader = new StreamReader(StructureFile);
}
// Search of structures one at a time
while (true)
{
if (fromList) // get next structure from list
{
if (listCidsRead >= cidList.Count)
{
break;
}
qid = cidList[listCidsRead].Cid;
listCidsRead++;
if (qid.Trim() == "")
{
continue;
}
if (qid.ToLower().IndexOf(".mol") > 0 || qid.ToLower().IndexOf(".skc") > 0)
{ // file reference
if (!File.Exists(qid))
{
continue;
}
if (qid.ToLower().IndexOf(".mol") > 0)
{
queryStructure = MoleculeMx.ReadMolfile(qid);
}
else
{
queryStructure = MoleculeMx.ReadSketchFile(qid);
}
}
else
{
queryStructure = MoleculeUtil.SelectMoleculeForCid(qid);
}
if (queryStructure == null || queryStructure.AtomCount == 0)
{
continue; // oops
}
}
else // get next structure from input file
{
qid = null;
if (StructureFile.ToLower().EndsWith(".sdf"))
{
List <SdFileField> fList = SdFileDao.Read(structureFileReader);
if (fList == null) // end of sdFile
{
structureFileReader.Close();
break;
}
if (fList.Count == 0)
{
continue;
}
queryStructure = new MoleculeMx(MoleculeFormat.Molfile, fList[0].Data);
if (queryStructure == null || queryStructure.AtomCount == 0)
{
continue;
}
if (keyField != "") // key field specified?
{
qid = SdFileDao.GetDataField(fList, keyField);
}
else // get name from 1st line of molfile
{
string molFile = fList[0].Data;
int i1 = molFile.IndexOf("\n");
if (i1 == 0)
{
qid = "";
}
else
{
qid = molFile.Substring(0, i1).Trim();
}
}
if (string.IsNullOrEmpty(qid))
{
qid = SdFileDao.GetDataField(fList, "compound_id");
}
}
else // assume smiles file
{
string smiles = structureFileReader.ReadLine();
if (smiles == null) // end of sdFile
{
structureFileReader.Close();
break;
}
smiles = smiles.Trim();
if (smiles.Length == 0)
{
continue;
}
int i1 = smiles.IndexOf(","); // get any preceeding queryId
if (i1 < 0)
{
i1 = smiles.IndexOf("\t");
}
if (i1 >= 0)
{
qid = smiles.Substring(0, i1).Trim();
smiles = smiles.Substring(i1 + 1).Trim();
}
queryStructure = new MoleculeMx(MoleculeFormat.Smiles, smiles);
if (queryStructure == null || queryStructure.AtomCount == 0)
{
continue;
}
}
if (qid == null || qid.Trim() == "")
{
qid = (queryCount + 1).ToString(); // be sure we have a query id
}
}
queryCount++; // count the query
if (queryStructure == null || queryStructure.AtomCount == 0)
{
WriteToLog("Error converting specific structure " + queryCount.ToString() + ", " + qid);
continue;
}
queryStructure.RemoveStructureCaption(); // remove any Mobius-added caption
Psc.Molecule = queryStructure;
string msg =
"Searching Structure: " + queryCount.ToString();
if (inputQueryCount > 0)
{
msg += " of " + inputQueryCount.ToString();
}
msg += "\n" +
"Structures with one or more matches: " + matchAtLeastOneCount.ToString() + "\n" +
"Total Matches: " + matchList.Count.ToString();
Progress.Show(msg);
// Do the search over the list of databases
for (int ti = 0; ti < tables.Count; ti++)
{
mt = tables[ti];
q = new Query(); // build basic query
//q.SingleStepExecution = true; // do in single step (doesn't currently return sim score)
q.ShowStereoComments = false;
qt = new QueryTable(mt);
q.AddQueryTable(qt);
qt.SelectKeyOnly(); // start selecting desired cols
keyMc = mt.KeyMetaColumn;
structMc = mt.FirstStructureMetaColumn;
structQc = qt.GetQueryColumnByName(structMc.Name);
structQc.Selected = RetrieveStructures;
dbSetMc = mt.DatabaseListMetaColumn;
if (dbSetMc == null)
{
throw new Exception("\"Databases\" metacolumn not found for table: " + mt.Label);
}
QueryColumn dbSetQc = qt.GetQueryColumnByName(dbSetMc.Name);
dbSetQc.Selected = true; // select the database name
RootTable root = RootTable.GetFromTableName(mt.Name);
txt = " in (" + root.Label + ")";
dbSetQc.Criteria = dbSetMc.Name + txt;
dbSetQc.CriteriaDisplay = txt;
simScoreMc = mt.SimilarityScoreMetaColumn;
simScoreQc = null;
if (simScoreMc != null) // get sim score if it exists
{
simScoreQc = qt.GetQueryColumnByName(simScoreMc.Name);
simScoreQc.Selected = true; // return sim score
}
Psc.QueryColumn = structQc;
ParsedStructureCriteria psc2 = AdjustSearchForSmallWorldAsNeeded(Psc);
psc2.ConvertToQueryColumnCriteria(structQc); // format the QC for the structure search
DateTime t0 = DateTime.Now;
//QueryEngine qe = new QueryEngine();
//qe.NextRowsMin = 1000; // minimum number of rows to prefetch
//qe.NextRowsMax = -1; // maximum number of rows to prefetch
//qe.NextRowsMaxTime = 10000; // max time in milliseconds for next fetch
//qe.ExecuteQuery(q);
qm = new QueryManager();
try { dlgRslt = qm.ExecuteQuery(ref q); }
catch (Exception ex)
{
WriteToLog("Error searching structure: " + ex.Message + ", " + queryCount.ToString() + ", " + qid);
continue;
}
if (dlgRslt != DialogResult.OK)
{
return(dlgRslt);
}
double executeTime = TimeOfDay.Delta(ref t0);
int offset = qm.DataTableManager.KeyValueVoPos + 1;
//int offset = 0; // for QE
int keyPos = offset++;
int strPos = RetrieveStructures ? offset++ : -1;
int dbPos = offset++;
int simPos = offset++;
int fetchCnt = 0;
while (true)
{
dr = qm.NextRow();
//dr = qe.NextRow(); // for Qe
if (dr == null)
{
break;
}
fetchCnt++;
if (fetchCnt == 1)
{
matchAtLeastOneCount++; // number of queries that have at least one match
}
MatchData md = new MatchData();
md.Qno = queryCount;
md.Qid = qid;
if (RetrieveStructures)
{
md.Qstr = "Chime=" + queryStructure.GetChimeString();
}
CompoundId cid = CompoundId.ConvertTo(dr[keyPos]);
md.Mid = cid.Value;
if (RetrieveStructures)
{
MoleculeMx ms = MoleculeMx.ConvertTo(dr[strPos]);
if (!NullValue.IsNull(ms))
{
md.Mstr = "Chime=" + ms.GetChimeString();
}
}
StringMx db = StringMx.ConvertTo(dr[dbPos]);
if (!NullValue.IsNull(db))
{
md.Db = db.Value;
}
if (Psc.SearchType == StructureSearchType.MolSim)
{
NumberMx nex = NumberMx.ConvertTo(dr[simPos]);
if (!NullValue.IsNull(nex))
{
md.Score = nex.Value;
}
}
if (matchList.Contains(cid.Value)) // already have compound id as match for other query?
{
if (Psc.SearchType != StructureSearchType.MolSim)
{
continue; // if similarity search see if more similar
}
CidListElement le = matchList.Get(cid.Value); // reference current score
if (le.Tag > md.Score)
{
continue; // only replace if more similar
}
matchList.Remove(le.Cid); // remove from list
for (int mi = 0; mi < matchData.Count; mi++) // remove from data
{
if (matchData[mi].Mid == md.Mid)
{
matchData.RemoveAt(mi);
break;
}
}
}
matchList.Add(md.Mid);
matchList.Get(md.Mid).Tag = md.Score; // keep score in list
matchData.Add(md); // add to results
txt = md.Qid + ", " + md.Mid + ", " + md.Score.ToString();
WriteToLog(txt);
} // Fetch result loop
double fetchTime = TimeOfDay.Delta(ref t0);
} // DB loop
if (Progress.CancelRequested)
{
Progress.Hide();
MessageBoxMx.ShowError("Search cancelled.");
try { LogStream.Close(); } catch { }
return(DialogResult.Cancel);
}
} // key loop
CidListCommand.WriteCurrentList(matchList); // write the list of numbers
UsageDao.LogEvent("MultipleStructSearch");
txt =
"=== Multiple structure search complete ===\r\n\r\n" +
"Structures Searched: " + queryCount.ToString() + "\r\n";
txt +=
"Structures with one or more matches: " + matchAtLeastOneCount.ToString() + "\r\n" +
"Total Matches: " + matchList.Count.ToString() + "\r\n";
TimeSpan ts = DateTime.Now.Subtract(startTime);
ts = new TimeSpan(ts.Hours, ts.Minutes, ts.Seconds);
txt += "Total Time: " + ts + "\r\n\r\n";
WriteToLog("\r\n" + txt);
try { LogStream.Close(); } catch { }
if (matchList.Count == 0)
{
Progress.Hide();
MessageBoxMx.ShowError("No matches have been found.");
return(DialogResult.Cancel);
}
tok = "Matching compound ids";
if (Psc.SearchType == StructureSearchType.MolSim)
{
tok = "Similar compound ids";
}
txt += tok + " have been written to the current list: " + ResultsUo.Name + "\n" +
"Log file written to: " + logFileName + "\n\n" +
"Do you want to view the match results?";
DialogResult dRslt = MessageBoxMx.Show(txt, "Multiple Structure Search", MessageBoxButtons.YesNoCancel, MessageBoxIcon.Question);
if (dRslt == DialogResult.Cancel)
{
return(DialogResult.Cancel);
}
else if (dRslt == DialogResult.No) // show log
{
SystemUtil.StartProcess(logFileName);
return(DialogResult.Cancel);
}
// Display results
Progress.Show("Formatting results...");
mt = new MetaTable();
mt.Name = "MULTSTRUCTSEARCH_" + SearchCount;
mt.Label = "Multiple Structure Search " + SearchCount;
mt.TableMap = "Mobius.Tools.MultStructSearch"; // plugin id
mt.MetaBrokerType = MetaBrokerType.NoSql;
ColumnSelectionEnum structureColumnSelection = RetrieveStructures ? ColumnSelectionEnum.Selected : ColumnSelectionEnum.Unselected;
keyMc = keyMc.Clone();
keyMc.Name = "MatchingCid";
keyMc.Label = "Matching Compound Id";
mt.AddMetaColumn(keyMc);
structMc = structMc.Clone();
structMc.Name = "MatchingStructure";
structMc.Label = "Matching Structure";
structMc.InitialSelection = structureColumnSelection;
mt.AddMetaColumn(structMc);
dbSetMc = dbSetMc.Clone();
dbSetMc.Name = "Database";
mt.AddMetaColumn(dbSetMc);
//if (DatabasesToSearch.Text.Contains(","))
dbSetMc.InitialSelection = ColumnSelectionEnum.Selected;
mc = mt.AddMetaColumn("Molsimilarity", "Similarity Search Score", MetaColumnType.Number, ColumnSelectionEnum.Unselected, 10);
if (Psc.SearchType == StructureSearchType.MolSim)
{
mc.InitialSelection = ColumnSelectionEnum.Selected;
}
mc = mt.AddMetaColumn("QueryNo", "Query Number", MetaColumnType.Integer);
//mc = mt.AddMetaColumn("QueryMatchNo", "Query Match Number", MetaColumnType.Integer);
mc = mt.AddMetaColumn("QueryId", "Query Id", MetaColumnType.String);
mc = mt.AddMetaColumn("QueryStructure", "Query Structure", MetaColumnType.Structure);
mc.InitialSelection = structureColumnSelection;
q = ToolHelper.InitEmbeddedDataToolQuery(mt);
dt = q.ResultsDataTable as DataTableMx;
for (int mi = 0; mi < matchData.Count; mi++)
{
MatchData md = matchData[mi];
dr = dt.NewRow();
dr[qt.Alias + ".MatchingCid"] = new CompoundId(md.Mid);
if (RetrieveStructures)
{
dr[qt.Alias + ".MatchingStructure"] = new MoleculeMx(MoleculeFormat.Chime, md.Mstr);
}
dr[qt.Alias + ".Database"] = new StringMx(md.Db);
if (Psc.SearchType == StructureSearchType.MolSim)
{
dr[qt.Alias + ".Molsimilarity"] = new NumberMx(md.Score);
}
dr[qt.Alias + ".QueryNo"] = new NumberMx(md.Qno);
dr[qt.Alias + ".QueryId"] = new StringMx(md.Qid);
if (RetrieveStructures)
{
dr[qt.Alias + ".QueryStructure"] = new MoleculeMx(MoleculeFormat.Chime, md.Qstr);
}
dt.Rows.Add(dr);
}
ToolHelper.DisplayData(q, dt, true);
Progress.Hide();
return(DialogResult.OK);
}
/// <summary>
/// OK button clicked, process input
/// </summary>
/// <returns></returns>
DialogResult ProcessInput()
{
int rgCount; // number of Rgroups
List <Rgroup> rgList; // list of existing Rgroups
bool[] rgExists; // entry = true if rgroup exists in core
Rgroup rg;
bool oneD, twoD; // matrix dimensionality
List <string> keys = null;
Dictionary <string, List <QualifiedNumber> > mElem; // matrix element dictionary
List <RgroupSubstituent>[] rSubs; // substituents seen for each Rgroup
Query q, q0, q2;
QueryTable qt, qt2;
QueryColumn qc, qc2;
MetaTable mt, mt2;
MetaColumn mc, mc2;
DataTableMx dt;
DataRowMx dr;
DialogResult dlgRslt;
string tok;
int ri, rii, si, qti, qci, bi, bi2;
// Get core structure & list of R-groups
MoleculeMx core = new MoleculeMx(MoleculeFormat.Molfile, SQuery.MolfileString);
if (core.AtomCount == 0)
{
MessageBoxMx.ShowError("A Core structure with R-groups must be defined");
return(DialogResult.None);
}
if (!Structure.Checked && !Smiles.Checked && !Formula.Checked &&
!Weight.Checked && !Index.Checked)
{
MessageBoxMx.ShowError("At least one substituent display format must be selected.");
return(DialogResult.None);
}
mt = MetaTableCollection.GetWithException("Rgroup_Decomposition");
qt = new QueryTable(mt);
qc = qt.GetQueryColumnByNameWithException("Core");
qc.MolString = core.GetMolfileString(); // put core structure into table criteria
qc.CriteriaDisplay = "Substructure search (SSS)";
qc.Criteria = "CORE SSS SQUERY";
qc = qt.GetQueryColumnByNameWithException("R1_Structure");
if (ShowCoreStructure.Checked)
{
qc.Label = "R-group, Core\tChime=" + core.GetChimeString(); // reference core in query col header label
qc.MetaColumn.Width = 25;
}
RgroupDecomposition.SetSelected(qt, "R1_Structure", Structure.Checked); // select for retrieval if checked
RgroupDecomposition.SetSelected(qt, "R1_Smiles", Smiles.Checked);
RgroupDecomposition.SetSelected(qt, "R1_Formula", Formula.Checked);
RgroupDecomposition.SetSelected(qt, "R1_Weight", Weight.Checked);
RgroupDecomposition.SetSelected(qt, "R1_SubstNo", Index.Checked);
string terminateOption = "First mapping"; // terminate on first complete match
qc = qt.GetQueryColumnByNameWithException("Terminate_Option");
qc.Criteria = qt.MetaTable.Name + " = " + Lex.AddSingleQuotes(terminateOption);
qc.CriteriaDisplay = "= " + Lex.AddSingleQuotes(terminateOption);
QueryTable rgdQt = qt; // keep a ref to it
if (QbUtil.Query == null || QbUtil.Query.Tables.Count == 0)
{
MessageBoxMx.ShowError("No current query.");
return(DialogResult.None);
}
q0 = QbUtil.Query; // original query this analysis is based on
q = q0.Clone(); // make copy of source query we can modify
q.SingleStepExecution = false;
qti = 0;
while (qti < q.Tables.Count) // deselect query columns that we don't want
{
qt = q.Tables[qti];
if (Lex.Eq(qt.MetaTable.Name, "Rgroup_Decomposition"))
{ // remove any rgroup decomp table
qti++;
continue;
}
mt = qt.MetaTable;
if (mt.MultiPivot || // check for tables not allowed in underlying query
mt.MetaBrokerType == MetaBrokerType.CalcField || // (called ShouldPresearchAndTransform previously)
mt.MetaBrokerType == MetaBrokerType.MultiTable ||
mt.MetaBrokerType == MetaBrokerType.RgroupDecomp)
{
MessageBoxMx.ShowError("Multipivot/Rgroup table \"" + qt.ActiveLabel +
"\" can't be included in an underlying Rgroup Matrix query");
return(DialogResult.None);
}
for (qci = 0; qci < qt.QueryColumns.Count; qci++)
{
qc = qt.QueryColumns[qci];
if (qc.MetaColumn == null)
{
continue;
}
switch (qc.MetaColumn.DataType)
{
case MetaColumnType.CompoundId: // keep only these
case MetaColumnType.Integer:
case MetaColumnType.Number:
case MetaColumnType.QualifiedNo:
case MetaColumnType.String:
break;
default:
qc.Selected = false;
break;
}
}
qti++;
}
q.AddQueryTable(rgdQt); // Add Rgroup decom table to end of cloned source query
Progress.Show("Retrieving data...");
try
{
dlgRslt = ToolHelper.ExecuteQuery(ref q, out keys);
if (dlgRslt != DialogResult.OK)
{
return(dlgRslt);
}
}
catch (Exception ex)
{
MessageBoxMx.ShowError("Error executing query:\r\n" + ex.Message);
return(DialogResult.None);
}
if (keys == null || keys.Count == 0)
{
Progress.Hide();
MessageBoxMx.ShowError("No results were returned by the query.");
return(DialogResult.None);
}
// Scan modified query to get list of rgroup indexes that are present
rgExists = new bool[32];
rgList = new List <Rgroup>();
QueryTable rgQt = q.GetQueryTableByName("Rgroup_Decomposition");
foreach (QueryColumn qc0 in rgQt.QueryColumns)
{
mc = qc0.MetaColumn;
if (!(mc.Name.StartsWith("R") && mc.Name.EndsWith("_STRUCTURE") && qc0.Selected))
{
continue; // skip if not a selected Rgroup structure
}
int len = mc.Name.Length - ("R" + "_STRUCTURE").Length;
tok = mc.Name.Substring(1, len);
if (!int.TryParse(tok, out ri))
{
continue;
}
rgExists[ri - 1] = true;
rg = new Rgroup();
rg.RIndex = ri;
rg.VoPos = qc0.VoPosition;
rgList.Add(rg);
}
for (bi = 1; bi < rgList.Count; bi++)
{ // sort by increasing R index
rg = rgList[bi];
for (bi2 = bi - 1; bi2 >= 0; bi2--)
{
if (rg.RIndex >= rgList[bi2].RIndex)
{
break;
}
rgList[bi2 + 1] = rgList[bi2];
}
rgList[bi2 + 1] = rg;
}
rgCount = rgList.Count;
twoD = TwoD.Checked;
if (rgCount == 1)
{
twoD = false; // if only 1 rgroup can't do as 2d
}
oneD = !twoD;
// Read data into mElem and rgroup substituents into rSubs.
// Matrix mElem is keyed on [R1Smiles, R2Smiles,... RnSmiles, FieldName] for 1d and
// [R1Smiles, R2Smiles,... FieldName, RnSmiles] for 2d
QueryManager qm = q.QueryManager as QueryManager;
DataTableManager dtm = qm.DataTableManager;
dt = qm.DataTable;
mElem = new Dictionary <string, List <QualifiedNumber> >(); // matrix element dictionary
rSubs = new List <RgroupSubstituent> [32]; // list of substituents seen for each Rgroup
for (rii = 0; rii < rgCount; rii++) // alloc substituent list for rgroup
{
rSubs[rii] = new List <RgroupSubstituent>();
}
int rowCount = 0;
while (true)
{ // scan data accumulating rgroup substituents and data values
dr = dtm.FetchNextDataRow();
if (dr == null)
{
break;
}
rowCount++;
string cid = dr[dtm.KeyValueVoPos] as string;
string lastMapCid = "", rgroupKey = "", rgroupKeyLast = "";
int mapCount = 0;
for (rii = 0; rii < rgCount; rii++) // for
{
MoleculeMx rSub = dr[rgList[rii].VoPos] as MoleculeMx;
if (rSub == null || rSub.AtomCount == 0)
{
continue;
}
ri = rgList[rii].RIndex; // actual R index in query
int subIdx = RgroupSubstituent.Get(rSub, rSubs[rii]);
// if (ri == 1 && subIdx != 0) subIdx = subIdx; // debug
if (subIdx < 0)
{
continue;
}
string rKey = "R" + ri.ToString() + "_" + (subIdx + 1).ToString();
if (oneD || rii < rgCount - 1)
{
if (rgroupKey != "")
{
rgroupKey += "\t";
}
rgroupKey += rKey;
}
else
{
rgroupKeyLast = rKey;
}
lastMapCid = cid;
mapCount++;
}
if (lastMapCid == cid) // add the data if compound has a mapping
{
AccumulateMatrixElements(mElem, q, dr, rgroupKey, rgroupKeyLast, cid);
}
if (Progress.IsTimeToUpdate)
{
Progress.Show("Retrieving data: " + StringMx.FormatIntegerWithCommas(rowCount) + " rows...");
}
}
if (rowCount == 0)
{
Progress.Hide();
MessageBoxMx.ShowError("No data rows retrieved");
return(DialogResult.None);
}
if (twoD && (rSubs[rgCount - 1] == null || rSubs[rgCount - 1].Count == 0))
{ // if 2D be sure we have at least one substituent for the last Rgroup
Progress.Hide();
MessageBoxMx.ShowError("No substituents found for R" + rgCount.ToString());
return(DialogResult.None);
}
// Create a MetaTable & DataTable for matrix results
Progress.Show("Analyzing data...");
mt = new MetaTable(); // create output table
MatrixCount++;
mt.Name = "RGROUPMATRIX_" + MatrixCount;
mt.Label = "R-group Matrix " + MatrixCount;
mt.MetaBrokerType = MetaBrokerType.RgroupDecomp;
mc = // use sequence for key
mt.AddMetaColumn("RgroupMatrixId", "No.", MetaColumnType.Integer, ColumnSelectionEnum.Selected, 3);
mc.ClickFunction = "None"; // avoid hyperlink on this key
mc.IsKey = true;
int maxLeftR = rgCount;
if (twoD)
{
maxLeftR = rgCount - 1;
}
for (ri = 0; ri < maxLeftR; ri++)
{
string rStr = "R" + (ri + 1).ToString();
if (Structure.Checked)
{
mc = mt.AddMetaColumn(rStr + "Str", rStr, MetaColumnType.Structure, ColumnSelectionEnum.Selected, 12);
if (ri == 0 && ShowCoreStructure.Checked) // include core structure above R1 if requested
{
string chimeString = MoleculeMx.MolfileStringToSmilesString(SQuery.MolfileString);
mc.Label = "R1, Core\tChime=" + chimeString;
mc.Width = 25;
}
}
if (Smiles.Checked)
{
mc = mt.AddMetaColumn(rStr + "Smi", rStr + " Smiles", MetaColumnType.String, ColumnSelectionEnum.Selected, 12);
}
if (Formula.Checked)
{
mc = mt.AddMetaColumn(rStr + "Mf", rStr + " Formula", MetaColumnType.String, ColumnSelectionEnum.Selected, 8);
}
if (Weight.Checked)
{
mc = mt.AddMetaColumn(rStr + "MW", rStr + " Mol. Wt.", MetaColumnType.Number, ColumnSelectionEnum.Selected, 6, ColumnFormatEnum.Decimal, 2);
}
if (Index.Checked)
{
mc = mt.AddMetaColumn(rStr + "Index", rStr + " Subst. Idx.", MetaColumnType.Number, ColumnSelectionEnum.Selected, 4);
mc.Format = ColumnFormatEnum.Decimal;
}
}
mc = // add column to contain result type
mt.AddMetaColumn("ResultType", "Result Type", MetaColumnType.String, ColumnSelectionEnum.Selected, 12);
if (oneD) // add just 1 column to contain results
{
mc = mt.AddMetaColumn("Results", "Results", MetaColumnType.QualifiedNo, ColumnSelectionEnum.Selected, 12);
mc.MetaBrokerType = MetaBrokerType.RgroupDecomp; // broker to do special col handling for cond formtting
if (QbUtil.Query.UserObject.Id > 0)
{
mc.DetailsAvailable = true;
}
}
else // add col for each substituent for last rgroup
{
string rStr = "R" + rgCount.ToString();
for (si = 0; si < rSubs[rgCount - 1].Count; si++)
{
string cName = rStr + "_" + (si + 1).ToString();
string cLabel = cName.Replace("_", ".");
RgroupSubstituent rgs = rSubs[ri][si]; // get substituent info
if (Structure.Checked) // include structure
{
cLabel += "\tChime=" + rgs.Struct.GetChimeString();
}
else if (Smiles.Checked)
{
cLabel += " = " + rgs.Struct.GetSmilesString();
}
else if (Formula.Checked)
{
cLabel += " = " + rgs.Struct.MolFormula;
}
else if (Weight.Checked)
{
cLabel += " = " + rgs.Struct.MolWeight;
}
else if (Index.Checked)
{
cLabel += " = " + (si + 1).ToString();
}
mc = mt.AddMetaColumn(cName, cLabel, MetaColumnType.QualifiedNo, ColumnSelectionEnum.Selected, 12);
mc.MetaBrokerType = MetaBrokerType.RgroupDecomp;
if (QbUtil.Query.UserObject.Id > 0)
{
mc.DetailsAvailable = true;
}
}
}
MetaTableCollection.UpdateGlobally(mt); // add as a known metatable
if (mElem.Count == 0) // be sure we have a matrix
{
Progress.Hide();
MessageBoxMx.ShowError("No matrix can be created because insufficient data was found.");
return(DialogResult.None);
}
// Build the DataTable
Progress.Show("Building data table...");
q2 = new Query(); // build single-table query to hold matrix
qt2 = new QueryTable(q2, mt);
dt = DataTableManager.BuildDataTable(q2);
Dictionary <string, List <QualifiedNumber> > .KeyCollection kc = mElem.Keys;
string[] rgKeys = new string[mElem.Count];
kc.CopyTo(rgKeys, 0);
Array.Sort(rgKeys);
string[] rgKey = null, lastRgKey = null;
int rki = 0;
for (rki = 0; rki < rgKeys.Length; rki++)
{
rgKey = rgKeys[rki].Split('\t');
int riTop = rgCount + 1; // all r substituents & field name on left
if (twoD)
{
riTop = rgCount;
}
for (ri = 0; ri < riTop; ri++) // see if any changes in left side substituents or field name
{
if (lastRgKey == null || rgKey[ri] != lastRgKey[ri])
{
break;
}
}
if (ri < riTop || oneD) // if 2d then new row only if some change before last R
{
dr = dt.NewRow();
dt.Rows.Add(dr);
dr[dtm.KeyValueVoPos + 1] = new NumberMx(dt.Rows.Count); // integer row key
}
if (!HideRepeatingSubstituents.Checked)
{
ri = 0; // start at first if not hiding
}
lastRgKey = rgKey;
for (ri = ri; ri < riTop; ri++) // build row with these
{
string rgSub = rgKey[ri]; // get substituent id or table.column name
if (rgSub == "")
{
continue;
}
if (ri < riTop - 1)
{ // output substituent and/or smiles
string rStr = "R" + (ri + 1).ToString();
si = rgSub.IndexOf("_");
si = Int32.Parse(rgSub.Substring(si + 1)) - 1; // get substituent index
RgroupSubstituent rgs = rSubs[ri][si]; // get substituent info
if (Structure.Checked)
{
qc2 = qt2.GetQueryColumnByName(rStr + "Str");
dr[QcToDcName(qc2)] = rgs.Struct;
}
if (Smiles.Checked)
{
qc2 = qt2.GetQueryColumnByName(rStr + "Smi");
dr[QcToDcName(qc2)] = new StringMx(rgs.Struct.GetSmilesString());
}
if (Formula.Checked)
{
qc2 = qt2.GetQueryColumnByName(rStr + "Mf");
dr[QcToDcName(qc2)] = new StringMx(rgs.Struct.MolFormula);
}
if (Weight.Checked)
{
qc2 = qt2.GetQueryColumnByName(rStr + "Mw");
dr[QcToDcName(qc2)] = new NumberMx(rgs.Struct.MolWeight);
}
if (Index.Checked)
{
qc2 = qt2.GetQueryColumnByName(rStr + "Index");
dr[QcToDcName(qc2)] = new NumberMx(si + 1);
}
}
else // output field name
{
string[] sa = rgSub.Split('.'); // get field name
qt = q.GetQueryTableByName(sa[0]);
qc = qt.GetQueryColumnByName(sa[1]);
string fieldName = qc.ActiveLabel;
if (q0.Tables.Count >= 3) // qualify by table if 3 or more tables in original query
{
fieldName = qt.ActiveLabel + " - " + fieldName;
}
qc2 = qt2.GetQueryColumnByName("ResultType");
dr[QcToDcName(qc2)] = new StringMx(fieldName);
}
}
// Output value
string cName;
if (oneD)
{
cName = "Results";
}
else
{
cName = rgKey[rgCount]; // get key for this substituent (e.g. R2_1)
}
if (Lex.IsUndefined(cName))
{
continue; // may be no substituent match
}
qc2 = qt2.GetQueryColumnByName(cName);
QualifiedNumber qn = SummarizeData(mElem[rgKeys[rki]]); // get summarized value
dr[QcToDcName(qc2)] = qn;
}
ToolHelper.DisplayData(q2, dt, true);
UsageDao.LogEvent("RgroupMatrix");
Progress.Hide();
return(DialogResult.OK);
}