Polished Cleaner/Converter classes, etc.

src/main/java/cpath/cleaner/HumanCycCleaner.java

@@ -103,12 +103,18 @@ protected void cleanXrefDBName(Model model)
 		for (Xref xr : model.getObjects(Xref.class))
 		{
 			if(xr.getDb() == null) {
-				if(!(xr instanceof PublicationXref)) 
+				if(!(xr instanceof PublicationXref))
 					LOG.warn(xr.getModelInterface().getSimpleName() + ".db is NULL; " + xr.getUri());
-			} else if(xr.getDb().startsWith("Entrez"))
+			}
+			else if(xr.getDb().startsWith("Entrez")) {
 				xr.setDb("genpept"); //Protein GenBank Identifier
-			else if(xr.getDb().equalsIgnoreCase("NCBI Taxonomy")) 
+			}
+			else if(xr.getDb().equalsIgnoreCase("NCBI Taxonomy")) {
 				xr.setDb("ncbitaxon");
+			}
+			else if(xr.getDb().equalsIgnoreCase("Ensembl Human")) {
+				xr.setDb("ensembl");
+			}
 		}
 	}
 

src/main/java/cpath/cleaner/ReactomeCleaner.java

@@ -30,36 +30,34 @@
  * Removes "unstable" Reactome ID xref from objects where a stable ID is present.
  */
 final class ReactomeCleaner implements Cleaner {
-    private static Logger log = LoggerFactory.getLogger(ReactomeCleaner.class);
+  private static Logger log = LoggerFactory.getLogger(ReactomeCleaner.class);
 
-    public void clean(InputStream data, OutputStream cleanedData)
+  public void clean(InputStream data, OutputStream cleanedData)
 	{	
 		// import the original Reactome BioPAX model from file
 		log.info("Cleaning Reactome data...");
 		SimpleIOHandler simpleReader = new SimpleIOHandler(BioPAXLevel.L3);
 		Model model = simpleReader.convertFromOWL(data);
-
 		// Normalize pathway URIs, where possible, using Reactome stable IDs
 		// Since v54, Reactome stable ID format has been changed to like: "R-HSA-123456"
 		final Map<String, Entity> newUriToEntityMap = new HashMap<>();
 		final Set<Process> processes = new HashSet<>(model.getObjects(Process.class));
 
-		for(Process proc : processes) {
-			if (StringUtils.contains(proc.getUri(),"identifiers.org/reactome")
-					|| StringUtils.contains(proc.getUri(), "bioregistry.io/reactome"))
+		for(Process proc : processes)
+		{
+			if (StringUtils.containsAny(proc.getUri(),
+					"identifiers.org/reactome", "bioregistry.io/reactome")) {
 				continue; //skip for already normalized pathway or interaction
-
-			final Set<UnificationXref> uxrefs = new ClassFilterSet<>(
-					new HashSet<>(proc.getXref()), UnificationXref.class);
+			}
+			final Set<UnificationXref> uxrefs = new ClassFilterSet<>(new HashSet<>(proc.getXref()), UnificationXref.class);
 			for (UnificationXref x : uxrefs) {
-				if (x.getDb() != null && x.getDb().equalsIgnoreCase("Reactome")) {
+				if (StringUtils.equalsIgnoreCase(x.getDb(),"reactome")) {
 					String stableId = x.getId();
-					//remove 'REACTOME:' (length=9) prefix if present (it's optional - according to MIRIAM)
+					//remove 'REACTOME:' (length=9) prefix if present (it's optional - according to MIRIAM/Bioregistry)
 					if (stableId.startsWith("REACTOME:")) {
 						stableId = stableId.substring(9);
 						// stableID is like 'R-HSA-123456'
 					}
-
 					final String uri = "bioregistry.io/reactome:" + stableId;
 					if (!model.containsID(uri) && !newUriToEntityMap.containsKey(uri)) {
 						//save it in the map to replace the URI later (see below)
@@ -79,9 +77,10 @@ public void clean(InputStream data, OutputStream cleanedData)
 			}
 		}
 
-		// set standard URIs for selected entities;
-		for(String uri : newUriToEntityMap.keySet())
+		// set standard URIs for selected entities (processes);
+		for(String uri : newUriToEntityMap.keySet()) {
 			CPathUtils.replaceUri(model, newUriToEntityMap.get(uri), uri);
+		}
 
 		// All Conversions in Reactome are LEFT-TO-RIGH, 
 		// unless otherwise was specified (confirmed with Guanming Wu, 2013/12)
@@ -91,29 +90,25 @@ public void clean(InputStream data, OutputStream cleanedData)
 				ent.setConversionDirection(ConversionDirectionType.LEFT_TO_RIGHT);
 		}
 
-		// Remove unstable UnificationXrefs like "Reactome Database ID Release XX"
+		// Remove unstable UnificationXrefs like "Reactome Database ID Release 65"
 		// if there is a stable xref in the same object
 		// Since Reactome v54, stable ID format is different (not like REACT_12345...)
 		final Set<Xref> xrefsToRemove = new HashSet<>();
-		for(Xref xref: new HashSet<>(model.getObjects(Xref.class))) {
-			if(xref.getDb() != null && xref.getDb().toLowerCase().startsWith("reactome database"))
+		for(Xref xref: new HashSet<>(model.getObjects(Xref.class)))
+		{
+			if(StringUtils.startsWithIgnoreCase(xref.getDb(),"reactome database"))
 			{
 				//remove the long comment (save some RAM)
-				if(!(xref instanceof PublicationXref))
+				if(!(xref instanceof PublicationXref)) {
 					xref.getComment().clear();
-
-				//proceed with a unification xref only...
-				if(xref instanceof UnificationXref) {
-					for(XReferrable owner :  new HashSet<>(xref.getXrefOf())) {
-						for(Xref x : new HashSet<>(owner.getXref())) {
-							if(!(x instanceof UnificationXref) || x.equals(xref))
-								continue;
-							//another unif. xref present in the same owner object
-							if(x.getDb() != null && x.getDb().equalsIgnoreCase("reactome")) {
-								//remove the unstable ID ref from the object that has a stable id
-								owner.removeXref(xref);
-								xrefsToRemove.add(xref);
-							}
+				}
+				for(XReferrable owner :  new HashSet<>(xref.getXrefOf())) {
+					for(Xref x : new HashSet<>(owner.getXref())) {
+						if(StringUtils.equalsIgnoreCase(x.getDb(), "reactome")) {
+							//if a standard "reactome" xref is also present in the same owner object,
+							//then remove the unstable ID xref from that object
+							owner.removeXref(xref);
+							xrefsToRemove.add(xref);
 						}
 					}
 				}
@@ -128,8 +123,7 @@ public void clean(InputStream data, OutputStream cleanedData)
 		try {
 			simpleReader.convertToOWL(model, cleanedData);
 		} catch (Exception e) {
-			throw new RuntimeException("clean(), Exception thrown while saving cleaned Reactome data", e);
+			throw new RuntimeException("clean(), failed saving the cleaned Reactome model", e);
 		}
 	}
-
 }

src/main/java/cpath/converter/UniprotConverter.java

@@ -70,13 +70,13 @@ public void convert(InputStream is, OutputStream os) {
 				// GN gene symbols - to PR names and rel. xrefs
 				if (geneName != null) {
 					Collection<String> geneNames = getGeneSymbols(geneName.toString(), proteinReference);
-					// always use "HGNC Symbol" for rel. xrefs, despite it can be from MGI, RGD,.. (these are coordinated by HGNC)
+					// always use HGNC Symbol for rel. xrefs, despite it can be from MGI, RGD (these are coordinated by HGNC);
 					// (cannot do this in setXRefsFromDRs: no gene synonyms there, and organism specific db names like MGI)
 					for (String symbol : geneNames) {
 						// also add Gene Names to PR names (can be >1 due to isoforms)
 						proteinReference.addName(symbol);
 						RelationshipXref rXRef = CPathUtils
-								.findOrCreateRelationshipXref(RelTypeVocab.IDENTITY, "HGNC Symbol", symbol, model);
+								.findOrCreateRelationshipXref(RelTypeVocab.IDENTITY, "hgnc.symbol", symbol, model);
 						proteinReference.addXref(rXRef);
 					}
 				}
@@ -305,12 +305,12 @@ else if (db.equalsIgnoreCase("EMBL")) {
 						fixedDb = "Nucleotide Sequence Database";
 						//last ID in a HGNC line is in fact gene name
 					} else if(db.equalsIgnoreCase("HGNC") && !id.startsWith("HGNC:")) {
-						fixedDb = "HGNC Symbol";
+						fixedDb = "hgnc.symbol";
 					}
 					//remove .version from RefSeq IDs
 					else if (db.equalsIgnoreCase("REFSEQ")) {
 						// extract only RefSeq AC from AC.Version ID form
-						fixedDb = "RefSeq";
+						fixedDb = "refseq";
 						id = id.replaceFirst("\\.\\d+", "");
 					}
 

src/main/java/cpath/service/BiopaxConverter.java

@@ -9,6 +9,7 @@
 
 import cpath.service.api.OutputFormat;
 import org.apache.commons.io.IOUtils;
+import org.apache.commons.lang3.StringUtils;
 import org.biopax.paxtools.io.gsea.GSEAConverter;
 import org.biopax.paxtools.io.jsonld.JsonldBiopaxConverter;
 import org.biopax.paxtools.io.jsonld.JsonldConverter;
@@ -178,9 +179,9 @@ private void convertToSBGN(Model m, OutputStream stream, Blacklist blackList, bo
   private void convertToGSEA(Model m, OutputStream stream, Map<String, String> options)
     throws IOException {
     String idType;
-    if ((idType = options.get("db")) == null)
-      idType = "hgnc symbol";
-
+    if ((idType = options.get("db")) == null) {
+      idType = "hgnc.symbol";
+    }
 
     // It won't traverse into sub-pathways; will use only pre-defined organisms.
     // GSEAConverter's 'skipSubPathways' option is a different beast from the PC web api's 'subpw':
@@ -215,7 +216,7 @@ private void convertToSIF(Model m, OutputStream out,
     ConfigurableIDFetcher idFetcher = new ConfigurableIDFetcher();
     idFetcher.chemDbStartsWithOrEquals("chebi");
 
-    if (db == null || db.isEmpty() || db.toLowerCase().startsWith("hgnc")) {
+    if (StringUtils.isBlank(db) || db.toLowerCase().startsWith("hgnc")) {
       idFetcher.seqDbStartsWithOrEquals("hgnc");
     } else if (db.toLowerCase().startsWith("uniprot")) {
       idFetcher.seqDbStartsWithOrEquals("uniprot");

src/main/java/cpath/service/ConsoleApplication.java

@@ -391,14 +391,14 @@ private void writeScriptCommands(String bpFilename, PrintWriter writer, boolean
     final String commaSepTaxonomyIds = String.join(",", service.settings().getOrganismTaxonomyIds());
     if (exportToGSEA) {
       writer.println(String.format("%s %s '%s' '%s' %s 2>&1 &", javaRunPaxtools, "toGSEA", bpFilename,
-        prefix + "hgnc.gmt", "'hgnc symbol' 'organisms=" + commaSepTaxonomyIds + "'"));//'hgnc symbol' - important
+        prefix + "hgnc.gmt", "'hgnc.symbol' 'organisms=" + commaSepTaxonomyIds + "'"));//'hgnc symbol' - important
       writer.println(String.format("%s %s '%s' '%s' %s 2>&1 &", javaRunPaxtools, "toGSEA", bpFilename,
         prefix + "uniprot.gmt", "'uniprot' 'organisms=" + commaSepTaxonomyIds + "'"));
       writer.println("wait"); //important
       writer.println("echo \"Done converting " + bpFilename + " to GSEA.\"");
     }
     writer.println(String.format("%s %s '%s' '%s' %s 2>&1 &", javaRunPaxtools, "toSIF", bpFilename,
-      prefix + "hgnc.txt", "seqDb=hgnc -extended -andSif exclude=neighbor_of"));//'hgnc symbol' or 'hgnc' does not matter
+      prefix + "hgnc.txt", "seqDb=hgnc -extended -andSif exclude=neighbor_of"));
     //UniProt ID based extended SIF files can be huge, take too long to generate; skip for now.
     writer.println("wait"); //important
     writer.println("echo \"Done converting " + bpFilename + " to SIF.\"");

src/main/java/cpath/service/Merger.java

@@ -606,8 +606,9 @@ else if(primaryACs.size() > maxNumXrefsToAdd) {
 		}
 
 		// map primary ACs to HGNC Symbols and generate RXs if not too many...
-		if (noneXrefDbStartsWith(bpe, "hgnc symbol"))
+		if (noneXrefDbStartsWith(bpe, "hgnc.symbol")) {
 			mayAddHgncXrefs(m, bpe, primaryACs, maxNumXrefsToAdd);
+		}
 	}
 
 	// For biopolymers, also map uniprot accessions to HGNC Symbols, and add the xrefs, if possible -
@@ -618,17 +619,17 @@ private void mayAddHgncXrefs(final Model m, final XReferrable bpe,
 		}
 		final Set<String> hgncSymbols = new HashSet<>();
 		for (String ac : accessions) {
-			ProteinReference canonicalPR =
-					(ProteinReference) warehouseModel.getByID("bioregistry.io/uniprot:" + ac);
+			ProteinReference canonicalPR = (ProteinReference) warehouseModel.getByID("bioregistry.io/uniprot:" + ac);
 			if (canonicalPR != null) {
 				for (Xref x : canonicalPR.getXref())
-					if (x.getDb().equalsIgnoreCase("hgnc symbol"))
+					if (x.getDb().equalsIgnoreCase("hgnc.symbol")) {
 						hgncSymbols.add(x.getId());
+					}
 			}
 		}
 		// add rel. xrefs if there are not too many (there's risk to make nonsense SIF/GSEA export...)
 		if (!hgncSymbols.isEmpty() && hgncSymbols.size() <= maxNumXrefsToAdd) {
-			addRelXrefs(m, bpe, "hgnc symbol", hgncSymbols, RelTypeVocab.ADDITIONAL_INFORMATION);
+			addRelXrefs(m, bpe, "hgnc.symbol", hgncSymbols, RelTypeVocab.ADDITIONAL_INFORMATION);
 		}
 	}