- public StockholmFile(String inFile, String type) throws IOException\r
- {\r
- super(inFile, type);\r
- }\r
-\r
- public StockholmFile(FileParse source) throws IOException\r
- {\r
- super(source);\r
- }\r
-\r
- public void initData()\r
- {\r
- super.initData();\r
- }\r
-\r
- /**\r
- * Parse a file in Stockholm format using VARNA\r
- */\r
- public void _parse_withVARNA(java.io.File inFile) throws IOException\r
- {\r
- FileReader fr = null;\r
- fr = new FileReader(inFile);\r
-\r
- BufferedReader r = new BufferedReader(fr);\r
- result = null;\r
- try\r
- {\r
- result = RNAFactory.loadSecStrStockholm(r);\r
- } catch (ExceptionUnmatchedClosingParentheses umcp)\r
- {\r
- errormessage = "Unmatched parentheses in annotation. Aborting ("\r
- + umcp.getMessage() + ")";\r
- throw new IOException(umcp);\r
- }\r
- // DEBUG System.out.println("this is the secondary scructure:"\r
- // +result.size());\r
- SequenceI[] seqs = new SequenceI[result.size()];\r
- for (int i = 0; i < result.size(); i++)\r
- {\r
- // DEBUG System.err.println("Processing i'th sequence in Stockholm file")\r
- RNA current = result.get(i);\r
-\r
- String seq = current.getSeq();\r
- String rna = current.getStructDBN(true);\r
- // DEBUG System.out.println(seq);\r
- // DEBUG System.err.println(rna);\r
- int begin = 0;\r
- int end = seq.length() - 1;\r
- id = safeName(getDataName());\r
- seqs[i] = new Sequence(id, seq, begin, end);\r
- String[] annot = new String[rna.length()];\r
- Annotation[] ann = new Annotation[rna.length()];\r
- for (int j = 0; j < rna.length(); j++)\r
- {\r
- annot[j] = rna.substring(j, j + 1);\r
-\r
- }\r
-\r
- for (int k = 0; k < rna.length(); k++)\r
- {\r
- ann[k] = new Annotation(annot[k], "",\r
- jalview.schemes.ResidueProperties.getRNASecStrucState(\r
- annot[k]).charAt(0), 0f);\r
-\r
- }\r
- AlignmentAnnotation align = new AlignmentAnnotation("Sec. str.",\r
- current.getID(), ann);\r
-\r
- seqs[i].addAlignmentAnnotation(align);\r
- seqs[i].setRNA(result.get(i));\r
- this.annotations.addElement(align);\r
- }\r
- this.setSeqs(seqs);\r
-\r
- }\r
-\r
- \r
- /**\r
- * Parse a file in Stockholm format into Jalview's data model. The file has to\r
- * be passed at construction time\r
- * \r
- * @throws IOException\r
- * If there is an error with the input file\r
- */\r
- public void parse() throws IOException\r
- {\r
- StringBuffer treeString = new StringBuffer();\r
- String treeName = null;\r
- // --------------- Variable Definitions -------------------\r
- String line;\r
- String version;\r
- // String id;\r
- Hashtable seqAnn = new Hashtable(); // Sequence related annotations\r
- Hashtable seqs = new Hashtable();\r
- Regex p, r, rend, s, x;\r
-\r
- // Temporary line for processing RNA annotation\r
- // String RNAannot = "";\r
-\r
- // ------------------ Parsing File ----------------------\r
- // First, we have to check that this file has STOCKHOLM format, i.e. the\r
- // first line must match\r
- r = new Regex("# STOCKHOLM ([\\d\\.]+)");\r
- if (!r.search(nextLine()))\r
- {\r
- throw new IOException(\r
- "This file is not in valid STOCKHOLM format: First line does not contain '# STOCKHOLM'");\r
- }\r
- else\r
- {\r
- version = r.stringMatched(1);\r
- // logger.debug("Stockholm version: " + version);\r
- }\r
-\r
- // We define some Regexes here that will be used regularily later\r
- rend = new Regex("^\\s*\\/\\/"); // Find the end of an alignment\r
- p = new Regex("(\\S+)\\/(\\d+)\\-(\\d+)"); // split sequence id in\r
- // id/from/to\r
- s = new Regex("(\\S+)\\s+(\\S*)\\s+(.*)"); // Parses annotation subtype\r
- r = new Regex("#=(G[FSRC]?)\\s+(.*)"); // Finds any annotation line\r
- x = new Regex("(\\S+)\\s+(\\S+)"); // split id from sequence\r
-\r
- // Convert all bracket types to parentheses (necessary for passing to VARNA)\r
- Regex openparen = new Regex("(<|\\[)", "(");\r
- Regex closeparen = new Regex("(>|\\])", ")");\r
-\r
- // Detect if file is RNA by looking for bracket types\r
- Regex detectbrackets = new Regex("(<|>|\\[|\\]|\\(|\\))");\r
-\r
- rend.optimize();\r
- p.optimize();\r
- s.optimize();\r
- r.optimize();\r
- x.optimize();\r
- openparen.optimize();\r
- closeparen.optimize();\r
-\r
- while ((line = nextLine()) != null)\r
- {\r
- if (line.length() == 0)\r
- {\r
- continue;\r
- }\r
- if (rend.search(line))\r
- {\r
- // End of the alignment, pass stuff back\r
- this.noSeqs = seqs.size();\r
-\r
- String seqdb,dbsource = null;\r
- Regex pf = new Regex("PF[0-9]{5}(.*)"); // Finds AC for Pfam\r
- Regex rf = new Regex("RF[0-9]{5}(.*)"); // Finds AC for Rfam\r
- if (getAlignmentProperty("AC") != null)\r
- {\r
- String dbType = getAlignmentProperty("AC").toString();\r
- if (pf.search(dbType))\r
- {\r
- // PFAM Alignment - so references are typically from Uniprot\r
- dbsource = "PFAM";\r
- }\r
- else if (rf.search(dbType))\r
- {\r
- dbsource = "RFAM";\r
- }\r
- }\r
- // logger.debug("Number of sequences: " + this.noSeqs);\r
- Enumeration accs = seqs.keys();\r
- while (accs.hasMoreElements())\r
- {\r
- String acc = (String) accs.nextElement();\r
- // logger.debug("Processing sequence " + acc);\r
- String seq = (String) seqs.remove(acc);\r
- if (maxLength < seq.length())\r
- {\r
- maxLength = seq.length();\r
- }\r
- int start = 1;\r
- int end = -1;\r
- String sid = acc;\r
- /*\r
- * Retrieve hash of annotations for this accession Associate\r
- * Annotation with accession\r
- */\r
- Hashtable accAnnotations = null;\r
-\r
- if (seqAnn != null && seqAnn.containsKey(acc))\r
- {\r
- accAnnotations = (Hashtable) seqAnn.remove(acc);\r
- // TODO: add structures to sequence\r
- }\r
-\r
- // Split accession in id and from/to\r
- if (p.search(acc))\r
- {\r
- sid = p.stringMatched(1);\r
- start = Integer.parseInt(p.stringMatched(2));\r
- end = Integer.parseInt(p.stringMatched(3));\r
- }\r
- // logger.debug(sid + ", " + start + ", " + end);\r
-\r
- Sequence seqO = new Sequence(sid, seq, start, end);\r
- // Add Description (if any)\r
- if (accAnnotations != null && accAnnotations.containsKey("DE"))\r
- {\r
- String desc = (String) accAnnotations.get("DE");\r
- seqO.setDescription((desc == null) ? "" : desc);\r
- }\r
-\r
- // Add DB References (if any)\r
- if (accAnnotations != null && accAnnotations.containsKey("DR"))\r
- {\r
- String dbr = (String) accAnnotations.get("DR");\r
- if (dbr != null && dbr.indexOf(";") > -1)\r
- {\r
- String src = dbr.substring(0, dbr.indexOf(";"));\r
- String acn = dbr.substring(dbr.indexOf(";") + 1);\r
- jalview.util.DBRefUtils.parseToDbRef(seqO, src, "0", acn);\r
- }\r
- }\r
-\r
- if (accAnnotations != null && accAnnotations.containsKey("AC"))\r
- {\r
- if (dbsource != null)\r
- {\r
- String dbr = (String) accAnnotations.get("AC");\r
- if (dbr != null)\r
- {\r
- // we could get very clever here - but for now - just try to guess accession type from source of alignment plus structure of accession\r
- guessDatabaseFor(seqO, dbr, dbsource);\r
- \r
- }\r
- } \r
- // else - do what ? add the data anyway and prompt the user to specify what references these are ?\r
- }\r
-\r
- Hashtable features = null;\r
- // We need to adjust the positions of all features to account for gaps\r
- try\r
- {\r
- features = (Hashtable) accAnnotations.remove("features");\r
- } catch (java.lang.NullPointerException e)\r
- {\r
- // loggerwarn("Getting Features for " + acc + ": " +\r
- // e.getMessage());\r
- // continue;\r
- }\r
- // if we have features\r
- if (features != null)\r
- {\r
- int posmap[] = seqO.findPositionMap();\r
- Enumeration i = features.keys();\r
- while (i.hasMoreElements())\r
- {\r
- // TODO: parse out secondary structure annotation as annotation\r
- // row\r
- // TODO: parse out scores as annotation row\r
- // TODO: map coding region to core jalview feature types\r
- String type = i.nextElement().toString();\r
- Hashtable content = (Hashtable) features.remove(type);\r
-\r
- // add alignment annotation for this feature\r
- String key = type2id(type);\r
- if (key != null)\r
- {\r
- if (accAnnotations != null\r
- && accAnnotations.containsKey(key))\r
- {\r
- Vector vv = (Vector) accAnnotations.get(key);\r
- for (int ii = 0; ii < vv.size(); ii++)\r
- {\r
- AlignmentAnnotation an = (AlignmentAnnotation) vv\r
- .elementAt(ii);\r
- seqO.addAlignmentAnnotation(an);\r
- }\r
- }\r
- }\r
-\r
- Enumeration j = content.keys();\r
- while (j.hasMoreElements())\r
- {\r
- String desc = j.nextElement().toString();\r
- String ns = content.get(desc).toString();\r
- char[] byChar = ns.toCharArray();\r
- for (int k = 0; k < byChar.length; k++)\r
- {\r
- char c = byChar[k];\r
- if (!(c == ' ' || c == '_' || c == '-' || c == '.')) // PFAM\r
- // uses\r
- // '.'\r
- // for\r
- // feature\r
- // background\r
- {\r
- int new_pos = posmap[k]; // look up nearest seqeunce\r
- // position to this column\r
- SequenceFeature feat = new SequenceFeature(type, desc,\r
- new_pos, new_pos, 0f, null);\r
-\r
- seqO.addSequenceFeature(feat);\r
- }\r
- }\r
- }\r
-\r
- }\r
-\r
- }\r
- // garbage collect\r
-\r
- // logger.debug("Adding seq " + acc + " from " + start + " to " + end\r
- // + ": " + seq);\r
- this.seqs.addElement(seqO);\r
- }\r
- return; // finished parsing this segment of source\r
- }\r
- else if (!r.search(line))\r
- {\r
- // System.err.println("Found sequence line: " + line);\r
-\r
- // Split sequence in sequence and accession parts\r
- if (!x.search(line))\r
- {\r
- // logger.error("Could not parse sequence line: " + line);\r
- throw new IOException("Could not parse sequence line: " + line);\r
- }\r
- String ns = (String) seqs.get(x.stringMatched(1));\r
- if (ns == null)\r
- {\r
- ns = "";\r
- }\r
- ns += x.stringMatched(2);\r
-\r
- seqs.put(x.stringMatched(1), ns);\r
- }\r
- else\r
- {\r
- String annType = r.stringMatched(1);\r
- String annContent = r.stringMatched(2);\r
-\r
- // System.err.println("type:" + annType + " content: " + annContent);\r
-\r
- if (annType.equals("GF"))\r
- {\r
- /*\r
- * Generic per-File annotation, free text Magic features: #=GF NH\r
- * <tree in New Hampshire eXtended format> #=GF TN <Unique identifier\r
- * for the next tree> Pfam descriptions: 7. DESCRIPTION OF FIELDS\r
- * \r
- * Compulsory fields: ------------------\r
- * \r
- * AC Accession number: Accession number in form PFxxxxx.version or\r
- * PBxxxxxx. ID Identification: One word name for family. DE\r
- * Definition: Short description of family. AU Author: Authors of the\r
- * entry. SE Source of seed: The source suggesting the seed members\r
- * belong to one family. GA Gathering method: Search threshold to\r
- * build the full alignment. TC Trusted Cutoff: Lowest sequence score\r
- * and domain score of match in the full alignment. NC Noise Cutoff:\r
- * Highest sequence score and domain score of match not in full\r
- * alignment. TP Type: Type of family -- presently Family, Domain,\r
- * Motif or Repeat. SQ Sequence: Number of sequences in alignment. AM\r
- * Alignment Method The order ls and fs hits are aligned to the model\r
- * to build the full align. // End of alignment.\r
- * \r
- * Optional fields: ----------------\r
- * \r
- * DC Database Comment: Comment about database reference. DR Database\r
- * Reference: Reference to external database. RC Reference Comment:\r
- * Comment about literature reference. RN Reference Number: Reference\r
- * Number. RM Reference Medline: Eight digit medline UI number. RT\r
- * Reference Title: Reference Title. RA Reference Author: Reference\r
- * Author RL Reference Location: Journal location. PI Previous\r
- * identifier: Record of all previous ID lines. KW Keywords: Keywords.\r
- * CC Comment: Comments. NE Pfam accession: Indicates a nested domain.\r
- * NL Location: Location of nested domains - sequence ID, start and\r
- * end of insert.\r
- * \r
- * Obsolete fields: ----------- AL Alignment method of seed: The\r
- * method used to align the seed members.\r
- */\r
- // Let's save the annotations, maybe we'll be able to do something\r
- // with them later...\r
- Regex an = new Regex("(\\w+)\\s*(.*)");\r
- if (an.search(annContent))\r
- {\r
- if (an.stringMatched(1).equals("NH"))\r
- {\r
- treeString.append(an.stringMatched(2));\r
- }\r
- else if (an.stringMatched(1).equals("TN"))\r
- {\r
- if (treeString.length() > 0)\r
- {\r
- if (treeName == null)\r
- {\r
- treeName = "Tree " + (getTreeCount() + 1);\r
- }\r
- addNewickTree(treeName, treeString.toString());\r
- }\r
- treeName = an.stringMatched(2);\r
- treeString = new StringBuffer();\r
- }\r
- setAlignmentProperty(an.stringMatched(1), an.stringMatched(2));\r
- }\r
- }\r
- else if (annType.equals("GS"))\r
- {\r
- // Generic per-Sequence annotation, free text\r
- /*\r
- * Pfam uses these features: Feature Description ---------------------\r
- * ----------- AC <accession> ACcession number DE <freetext>\r
- * DEscription DR <db>; <accession>; Database Reference OS <organism>\r
- * OrganiSm (species) OC <clade> Organism Classification (clade, etc.)\r
- * LO <look> Look (Color, etc.)\r
- */\r
- if (s.search(annContent))\r
- {\r
- String acc = s.stringMatched(1);\r
- String type = s.stringMatched(2);\r
- String content = s.stringMatched(3);\r
- // TODO: store DR in a vector.\r
- // TODO: store AC according to generic file db annotation.\r
- Hashtable ann;\r
- if (seqAnn.containsKey(acc))\r
- {\r
- ann = (Hashtable) seqAnn.get(acc);\r
- }\r
- else\r
- {\r
- ann = new Hashtable();\r
- }\r
- ann.put(type, content);\r
- seqAnn.put(acc, ann);\r
- }\r
- else\r
- {\r
- throw new IOException("Error parsing " + line);\r
- }\r
- }\r
- else if (annType.equals("GC"))\r
- {\r
- // Generic per-Column annotation, exactly 1 char per column\r
- // always need a label.\r
- if (x.search(annContent))\r
- {\r
- // parse out and create alignment annotation directly.\r
- parseAnnotationRow(annotations, x.stringMatched(1),\r
- x.stringMatched(2));\r
- }\r
- }\r
- else if (annType.equals("GR"))\r
- {\r
- // Generic per-Sequence AND per-Column markup, exactly 1 char per\r
- // column\r
- /*\r
- * Feature Description Markup letters ------- -----------\r
- * -------------- SS Secondary Structure [HGIEBTSCX] SA Surface\r
- * Accessibility [0-9X] (0=0%-10%; ...; 9=90%-100%) TM TransMembrane\r
- * [Mio] PP Posterior Probability [0-9*] (0=0.00-0.05; 1=0.05-0.15;\r
- * *=0.95-1.00) LI LIgand binding [*] AS Active Site [*] IN INtron (in\r
- * or after) [0-2]\r
- */\r
- if (s.search(annContent))\r
- {\r
- String acc = s.stringMatched(1);\r
- String type = s.stringMatched(2);\r
- String seq = new String(s.stringMatched(3));\r
- String description = null;\r
- // Check for additional information about the current annotation\r
- // We use a simple string tokenizer here for speed\r
- StringTokenizer sep = new StringTokenizer(seq, " \t");\r
- description = sep.nextToken();\r
- if (sep.hasMoreTokens())\r
- {\r
- seq = sep.nextToken();\r
- }\r
- else\r
- {\r
- seq = description;\r
- description = new String();\r
- }\r
- // sequence id with from-to fields\r
-\r
- Hashtable ann;\r
- // Get an object with all the annotations for this sequence\r
- if (seqAnn.containsKey(acc))\r
- {\r
- // logger.debug("Found annotations for " + acc);\r
- ann = (Hashtable) seqAnn.get(acc);\r
- }\r
- else\r
- {\r
- // logger.debug("Creating new annotations holder for " + acc);\r
- ann = new Hashtable();\r
- seqAnn.put(acc, ann);\r
- }\r
- // TODO test structure, call parseAnnotationRow with vector from\r
- // hashtable for specific sequence\r
- Hashtable features;\r
- // Get an object with all the content for an annotation\r
- if (ann.containsKey("features"))\r
- {\r
- // logger.debug("Found features for " + acc);\r
- features = (Hashtable) ann.get("features");\r
- }\r
- else\r
- {\r
- // logger.debug("Creating new features holder for " + acc);\r
- features = new Hashtable();\r
- ann.put("features", features);\r
- }\r
-\r
- Hashtable content;\r
- if (features.containsKey(this.id2type(type)))\r
- {\r
- // logger.debug("Found content for " + this.id2type(type));\r
- content = (Hashtable) features.get(this.id2type(type));\r
- }\r
- else\r
- {\r
- // logger.debug("Creating new content holder for " +\r
- // this.id2type(type));\r
- content = new Hashtable();\r
- features.put(this.id2type(type), content);\r
- }\r
- String ns = (String) content.get(description);\r
- if (ns == null)\r
- {\r
- ns = "";\r
- }\r
- ns += seq;\r
- content.put(description, ns);\r
- Hashtable strucAnn;\r
- if (seqAnn.containsKey(acc))\r
- {\r
- strucAnn = (Hashtable) seqAnn.get(acc);\r
- }\r
- else\r
- {\r
- strucAnn = new Hashtable();\r
- }\r
-\r
- Vector newStruc = new Vector();\r
- parseAnnotationRow(newStruc, type, ns);\r
- strucAnn.put(type, newStruc);\r
- seqAnn.put(acc, strucAnn);\r
- }\r
- else\r
- {\r
- System.err\r
- .println("Warning - couldn't parse sequence annotation row line:\n"\r
- + line);\r
- // throw new IOException("Error parsing " + line);\r
- }\r
- }\r
- else\r
- {\r
- throw new IOException("Unknown annotation detected: " + annType\r
- + " " + annContent);\r
- }\r
- }\r
- }\r
- if (treeString.length() > 0)\r
- {\r
- if (treeName == null)\r
- {\r
- treeName = "Tree " + (1 + getTreeCount());\r
- }\r
- addNewickTree(treeName, treeString.toString());\r
- }\r
- }\r
-\r
- /**\r
- * Demangle an accession string and guess the originating sequence database for a given sequence\r
- * @param seqO sequence to be annotated\r
- * @param dbr Accession string for sequence\r
- * @param dbsource source database for alignment (PFAM or RFAM)\r
- */\r
- private void guessDatabaseFor(Sequence seqO, String dbr, String dbsource)\r
- {\r
- DBRefEntry dbrf=null;\r
- List<DBRefEntry> dbrs=new ArrayList<DBRefEntry>();\r
- String seqdb="Unknown",sdbac=""+dbr;\r
- int st=-1,en=-1,p;\r
- if ((st=sdbac.indexOf("/"))>-1)\r
- {\r
- String num,range=sdbac.substring(st+1);\r
- sdbac = sdbac.substring(0,st);\r
- if ((p=range.indexOf("-"))>-1)\r
- {\r
- p++;\r
- if (p<range.length())\r
- {\r
- num = range.substring(p).trim();\r
- try {\r
- en = Integer.parseInt(num);\r
- } catch (NumberFormatException x)\r
- {\r
- // could warn here that index is invalid\r
- en = -1;\r
- }\r
- }\r
- } else {\r
- p=range.length();\r
- }\r
- num=range.substring(0,p).trim();\r
- try {\r
- st = Integer.parseInt(num);\r
- } catch (NumberFormatException x)\r
- {\r
- // could warn here that index is invalid\r
- st = -1;\r
- }\r
- }\r
- if (dbsource.equals("PFAM")) {\r
- seqdb = "UNIPROT";\r
- if (sdbac.indexOf(".")>-1)\r
- {\r
- // strip of last subdomain\r
- sdbac = sdbac.substring(0,sdbac.indexOf("."));\r
- dbrf = jalview.util.DBRefUtils.parseToDbRef(seqO, seqdb, dbsource, sdbac);\r
- if (dbrf!=null)\r
- {\r
- dbrs.add(dbrf);\r
- }\r
- }\r
- dbrf = jalview.util.DBRefUtils.parseToDbRef(seqO, dbsource, dbsource, dbr);\r
- if (dbr!=null)\r
- {\r
- dbrs.add(dbrf);\r
- }\r
- } else {\r
- seqdb = "EMBL"; // total guess - could be ENA, or something else these days\r
- if (sdbac.indexOf(".")>-1)\r
- {\r
- // strip off last subdomain\r
- sdbac = sdbac.substring(0,sdbac.indexOf("."));\r
- dbrf = jalview.util.DBRefUtils.parseToDbRef(seqO, seqdb, dbsource, sdbac);\r
- if (dbrf!=null)\r
- {\r
- dbrs.add(dbrf);\r
- }\r
- }\r
- \r
- dbrf = jalview.util.DBRefUtils.parseToDbRef(seqO, dbsource, dbsource, dbr);\r
- if (dbrf!=null)\r
- {\r
- dbrs.add(dbrf);\r
- }\r
- }\r
- if (st!=-1 && en!=-1)\r
- {\r
- for (DBRefEntry d:dbrs)\r
- {\r
- jalview.util.MapList mp = new jalview.util.MapList(new int[] { seqO.getStart(),seqO.getEnd()},new int[] { st,en},1,1);\r
- jalview.datamodel.Mapping mping = new jalview.datamodel.Mapping(mp);\r
- d.setMap(mping);\r
- }\r
- }\r
- }\r
- \r
- protected static AlignmentAnnotation parseAnnotationRow(\r
- Vector annotation, String label, String annots)\r
- {\r
- String convert1, convert2 = null;\r
-\r
- // Convert all bracket types to parentheses\r
- Regex openparen = new Regex("(<|\\[)", "(");\r
- Regex closeparen = new Regex("(>|\\])", ")");\r
-\r
- // Detect if file is RNA by looking for bracket types\r
- Regex detectbrackets = new Regex("(<|>|\\[|\\]|\\(|\\))");\r
-\r
- convert1 = openparen.replaceAll(annots);\r
- convert2 = closeparen.replaceAll(convert1);\r
- annots = convert2;\r
-\r
+ public StockholmFile(String inFile, DataSourceType type)
+ throws IOException
+ {
+ super(inFile, type);
+ }
+
+ public StockholmFile(FileParse source) throws IOException
+ {
+ super(source);
+ }
+
+ @Override
+ public void initData()
+ {
+ super.initData();
+ }
+
+ /**
+ * Parse a file in Stockholm format into Jalview's data model using VARNA
+ *
+ * @throws IOException
+ * If there is an error with the input file
+ */
+ public void parse_with_VARNA(java.io.File inFile) throws IOException
+ {
+ FileReader fr = null;
+ fr = new FileReader(inFile);
+
+ BufferedReader r = new BufferedReader(fr);
+ List<RNA> result = null;
+ try
+ {
+ result = RNAFactory.loadSecStrStockholm(r);
+ } catch (ExceptionUnmatchedClosingParentheses umcp)
+ {
+ errormessage = "Unmatched parentheses in annotation. Aborting ("
+ + umcp.getMessage() + ")";
+ throw new IOException(umcp);
+ }
+ // DEBUG jalview.bin.Console.outPrintln("this is the secondary scructure:"
+ // +result.size());
+ SequenceI[] seqs = new SequenceI[result.size()];
+ String id = null;
+ for (int i = 0; i < result.size(); i++)
+ {
+ // DEBUG jalview.bin.Console.errPrintln("Processing i'th sequence in Stockholm file")
+ RNA current = result.get(i);
+
+ String seq = current.getSeq();
+ String rna = current.getStructDBN(true);
+ // DEBUG jalview.bin.Console.outPrintln(seq);
+ // DEBUG jalview.bin.Console.errPrintln(rna);
+ int begin = 0;
+ int end = seq.length() - 1;
+ id = safeName(getDataName());
+ seqs[i] = new Sequence(id, seq, begin, end);
+ String[] annot = new String[rna.length()];
+ Annotation[] ann = new Annotation[rna.length()];
+ for (int j = 0; j < rna.length(); j++)
+ {
+ annot[j] = rna.substring(j, j + 1);
+
+ }
+
+ for (int k = 0; k < rna.length(); k++)
+ {
+ ann[k] = new Annotation(annot[k], "",
+ Rna.getRNASecStrucState(annot[k]).charAt(0), 0f);
+
+ }
+ AlignmentAnnotation align = new AlignmentAnnotation("Sec. str.",
+ current.getID(), ann);
+
+ seqs[i].addAlignmentAnnotation(align);
+ seqs[i].setRNA(result.get(i));
+ this.annotations.addElement(align);
+ }
+ this.setSeqs(seqs);
+
+ }
+
+ /**
+ * Parse a file in Stockholm format into Jalview's data model. The file has to
+ * be passed at construction time
+ *
+ * @throws IOException
+ * If there is an error with the input file
+ */
+ @Override
+ public void parse() throws IOException
+ {
+ StringBuffer treeString = new StringBuffer();
+ String treeName = null;
+ // --------------- Variable Definitions -------------------
+ String line;
+ String version;
+ // String id;
+ Hashtable seqAnn = new Hashtable(); // Sequence related annotations
+ LinkedHashMap<String, String> seqs = new LinkedHashMap<>();
+ Regex p, r, rend, s, x;
+ // Temporary line for processing RNA annotation
+ // String RNAannot = "";
+
+ // ------------------ Parsing File ----------------------
+ // First, we have to check that this file has STOCKHOLM format, i.e. the
+ // first line must match
+
+ r = new Regex("# STOCKHOLM ([\\d\\.]+)");
+ if (!r.search(nextLine()))
+ {
+ throw new IOException(
+ MessageManager.getString("exception.stockholm_invalid_format")
+ + " (" + r + ")");
+ }
+ else
+ {
+ version = r.stringMatched(1);
+
+ // logger.debug("Stockholm version: " + version);
+ }
+
+ // We define some Regexes here that will be used regularily later
+ rend = new Regex("^\\s*\\/\\/"); // Find the end of an alignment
+ p = new Regex("(\\S+)\\/(\\d+)\\-(\\d+)"); // split sequence id in
+ // id/from/to
+ s = new Regex("(\\S+)\\s+(\\S*)\\s+(.*)"); // Parses annotation subtype
+ r = new Regex("#=(G[FSRC]?)\\s+(.*)"); // Finds any annotation line
+ x = new Regex("(\\S+)\\s+(\\S+)"); // split id from sequence
+
+ // Convert all bracket types to parentheses (necessary for passing to VARNA)
+ Regex openparen = new Regex("(<|\\[)", "(");
+ Regex closeparen = new Regex("(>|\\])", ")");
+
+ // // Detect if file is RNA by looking for bracket types
+ // Regex detectbrackets = new Regex("(<|>|\\[|\\]|\\(|\\))");
+
+ rend.optimize();
+ p.optimize();
+ s.optimize();
+ r.optimize();
+ x.optimize();
+ openparen.optimize();
+ closeparen.optimize();
+
+ while ((line = nextLine()) != null)
+ {
+ if (line.length() == 0)
+ {
+ continue;
+ }
+ if (rend.search(line))
+ {
+ // End of the alignment, pass stuff back
+ this.noSeqs = seqs.size();
+
+ String dbsource = null;
+ Regex pf = new Regex("PF[0-9]{5}(.*)"); // Finds AC for Pfam
+ Regex rf = new Regex("RF[0-9]{5}(.*)"); // Finds AC for Rfam
+ if (getAlignmentProperty("AC") != null)
+ {
+ String dbType = getAlignmentProperty("AC").toString();
+ if (pf.search(dbType))
+ {
+ // PFAM Alignment - so references are typically from Uniprot
+ dbsource = "PFAM";
+ }
+ else if (rf.search(dbType))
+ {
+ dbsource = "RFAM";
+ }
+ }
+ // logger.debug("Number of sequences: " + this.noSeqs);
+ for (Map.Entry<String, String> skey : seqs.entrySet())
+ {
+ // logger.debug("Processing sequence " + acc);
+ String acc = skey.getKey();
+ String seq = skey.getValue();
+ if (maxLength < seq.length())
+ {
+ maxLength = seq.length();
+ }
+ int start = 1;
+ int end = -1;
+ String sid = acc;
+ /*
+ * Retrieve hash of annotations for this accession Associate
+ * Annotation with accession
+ */
+ Hashtable accAnnotations = null;
+
+ if (seqAnn != null && seqAnn.containsKey(acc))
+ {
+ accAnnotations = (Hashtable) seqAnn.remove(acc);
+ // TODO: add structures to sequence
+ }
+
+ // Split accession in id and from/to
+ if (p.search(acc))
+ {
+ sid = p.stringMatched(1);
+ start = Integer.parseInt(p.stringMatched(2));
+ end = Integer.parseInt(p.stringMatched(3));
+ }
+ // logger.debug(sid + ", " + start + ", " + end);
+
+ Sequence seqO = new Sequence(sid, seq, start, end);
+ // Add Description (if any)
+ if (accAnnotations != null && accAnnotations.containsKey("DE"))
+ {
+ String desc = (String) accAnnotations.get("DE");
+ seqO.setDescription((desc == null) ? "" : desc);
+ }
+ // Add DB References (if any)
+ if (accAnnotations != null && accAnnotations.containsKey("DR"))
+ {
+ String dbr = (String) accAnnotations.get("DR");
+ if (dbr != null && dbr.indexOf(";") > -1)
+ {
+ String src = dbr.substring(0, dbr.indexOf(";"));
+ String acn = dbr.substring(dbr.indexOf(";") + 1);
+ jalview.util.DBRefUtils.parseToDbRef(seqO, src, "0", acn);
+ }
+ }
+
+ if (accAnnotations != null && accAnnotations.containsKey("AC"))
+ {
+ String dbr = (String) accAnnotations.get("AC");
+ if (dbr != null)
+ {
+ // we could get very clever here - but for now - just try to
+ // guess accession type from type of sequence, source of alignment
+ // plus
+ // structure
+ // of accession
+ guessDatabaseFor(seqO, dbr, dbsource);
+ }
+ // else - do what ? add the data anyway and prompt the user to
+ // specify what references these are ?
+ }
+
+ Hashtable features = null;
+ // We need to adjust the positions of all features to account for gaps
+ try
+ {
+ features = (Hashtable) accAnnotations.remove("features");
+ } catch (java.lang.NullPointerException e)
+ {
+ // loggerwarn("Getting Features for " + acc + ": " +
+ // e.getMessage());
+ // continue;
+ }
+ // if we have features
+ if (features != null)
+ {
+ int posmap[] = seqO.findPositionMap();
+ Enumeration i = features.keys();
+ while (i.hasMoreElements())
+ {
+ // TODO: parse out secondary structure annotation as annotation
+ // row
+ // TODO: parse out scores as annotation row
+ // TODO: map coding region to core jalview feature types
+ String type = i.nextElement().toString();
+ Hashtable content = (Hashtable) features.remove(type);
+
+ // add alignment annotation for this feature
+ String key = type2id(type);
+
+ /*
+ * have we added annotation rows for this type ?
+ */
+ boolean annotsAdded = false;
+ if (key != null)
+ {
+ if (accAnnotations != null
+ && accAnnotations.containsKey(key))
+ {
+ Vector vv = (Vector) accAnnotations.get(key);
+ for (int ii = 0; ii < vv.size(); ii++)
+ {
+ annotsAdded = true;
+ AlignmentAnnotation an = (AlignmentAnnotation) vv
+ .elementAt(ii);
+ seqO.addAlignmentAnnotation(an);
+ annotations.add(an);
+ }
+ }
+ }
+
+ Enumeration j = content.keys();
+ while (j.hasMoreElements())
+ {
+ String desc = j.nextElement().toString();
+ if (ANNOTATION.equals(desc) && annotsAdded)
+ {
+ // don't add features if we already added an annotation row
+ continue;
+ }
+ String ns = content.get(desc).toString();
+ char[] byChar = ns.toCharArray();
+ for (int k = 0; k < byChar.length; k++)
+ {
+ char c = byChar[k];
+ if (!(c == ' ' || c == '_' || c == '-' || c == '.')) // PFAM
+ // uses
+ // '.'
+ // for
+ // feature
+ // background
+ {
+ int new_pos = posmap[k]; // look up nearest seqeunce
+ // position to this column
+ SequenceFeature feat = new SequenceFeature(type, desc,
+ new_pos, new_pos, null);
+
+ seqO.addSequenceFeature(feat);
+ }
+ }
+ }
+
+ }
+
+ }
+ // garbage collect
+
+ // logger.debug("Adding seq " + acc + " from " + start + " to " + end
+ // + ": " + seq);
+ this.seqs.addElement(seqO);
+ }
+ return; // finished parsing this segment of source
+ }
+ else if (!r.search(line))
+ {
+ // jalview.bin.Console.errPrintln("Found sequence line: " + line);
+
+ // Split sequence in sequence and accession parts
+ if (!x.search(line))
+ {
+ // logger.error("Could not parse sequence line: " + line);
+ throw new IOException(MessageManager.formatMessage(
+ "exception.couldnt_parse_sequence_line", new String[]
+ { line }));
+ }
+ String ns = seqs.get(x.stringMatched(1));
+ if (ns == null)
+ {
+ ns = "";
+ }
+ ns += x.stringMatched(2);
+
+ seqs.put(x.stringMatched(1), ns);
+ }
+ else
+ {
+ String annType = r.stringMatched(1);
+ String annContent = r.stringMatched(2);
+
+ // jalview.bin.Console.errPrintln("type:" + annType + " content: " + annContent);
+
+ if (annType.equals("GF"))
+ {
+ /*
+ * Generic per-File annotation, free text Magic features: #=GF NH
+ * <tree in New Hampshire eXtended format> #=GF TN <Unique identifier
+ * for the next tree> Pfam descriptions: 7. DESCRIPTION OF FIELDS
+ *
+ * Compulsory fields: ------------------
+ *
+ * AC Accession number: Accession number in form PFxxxxx.version or
+ * PBxxxxxx. ID Identification: One word name for family. DE
+ * Definition: Short description of family. AU Author: Authors of the
+ * entry. SE Source of seed: The source suggesting the seed members
+ * belong to one family. GA Gathering method: Search threshold to
+ * build the full alignment. TC Trusted Cutoff: Lowest sequence score
+ * and domain score of match in the full alignment. NC Noise Cutoff:
+ * Highest sequence score and domain score of match not in full
+ * alignment. TP Type: Type of family -- presently Family, Domain,
+ * Motif or Repeat. SQ Sequence: Number of sequences in alignment. AM
+ * Alignment Method The order ls and fs hits are aligned to the model
+ * to build the full align. // End of alignment.
+ *
+ * Optional fields: ----------------
+ *
+ * DC Database Comment: Comment about database reference. DR Database
+ * Reference: Reference to external database. RC Reference Comment:
+ * Comment about literature reference. RN Reference Number: Reference
+ * Number. RM Reference Medline: Eight digit medline UI number. RT
+ * Reference Title: Reference Title. RA Reference Author: Reference
+ * Author RL Reference Location: Journal location. PI Previous
+ * identifier: Record of all previous ID lines. KW Keywords: Keywords.
+ * CC Comment: Comments. NE Pfam accession: Indicates a nested domain.
+ * NL Location: Location of nested domains - sequence ID, start and
+ * end of insert.
+ *
+ * Obsolete fields: ----------- AL Alignment method of seed: The
+ * method used to align the seed members.
+ */
+ // Let's save the annotations, maybe we'll be able to do something
+ // with them later...
+ Regex an = new Regex("(\\w+)\\s*(.*)");
+ if (an.search(annContent))
+ {
+ if (an.stringMatched(1).equals("NH"))
+ {
+ treeString.append(an.stringMatched(2));
+ }
+ else if (an.stringMatched(1).equals("TN"))
+ {
+ if (treeString.length() > 0)
+ {
+ if (treeName == null)
+ {
+ treeName = "Tree " + (getTreeCount() + 1);
+ }
+ addNewickTree(treeName, treeString.toString());
+ }
+ treeName = an.stringMatched(2);
+ treeString = new StringBuffer();
+ }
+ // TODO: JAL-3532 - this is where GF comments and database
+ // references are lost
+ // suggest overriding this method for Stockholm files to catch and
+ // properly
+ // process CC, DR etc into multivalued properties
+ setAlignmentProperty(an.stringMatched(1), an.stringMatched(2));
+ }
+ }
+ else if (annType.equals("GS"))
+ {
+ // Generic per-Sequence annotation, free text
+ /*
+ * Pfam uses these features: Feature Description ---------------------
+ * ----------- AC <accession> ACcession number DE <freetext>
+ * DEscription DR <db>; <accession>; Database Reference OS <organism>
+ * OrganiSm (species) OC <clade> Organism Classification (clade, etc.)
+ * LO <look> Look (Color, etc.)
+ */
+ if (s.search(annContent))
+ {
+ String acc = s.stringMatched(1);
+ String type = s.stringMatched(2);
+ String content = s.stringMatched(3);
+ // TODO: store DR in a vector.
+ // TODO: store AC according to generic file db annotation.
+ Hashtable ann;
+ if (seqAnn.containsKey(acc))
+ {
+ ann = (Hashtable) seqAnn.get(acc);
+ }
+ else
+ {
+ ann = new Hashtable();
+ }
+ ann.put(type, content);
+ seqAnn.put(acc, ann);
+ }
+ else
+ {
+ // throw new IOException("Error parsing " + line);
+ jalview.bin.Console.errPrintln(">> missing annotation: " + line);
+ }
+ }
+ else if (annType.equals("GC"))
+ {
+ // Generic per-Column annotation, exactly 1 char per column
+ // always need a label.
+ if (x.search(annContent))
+ {
+ // parse out and create alignment annotation directly.
+ parseAnnotationRow(annotations, x.stringMatched(1),
+ x.stringMatched(2));
+ }
+ }
+ else if (annType.equals("GR"))
+ {
+ // Generic per-Sequence AND per-Column markup, exactly 1 char per
+ // column
+ /*
+ * Feature Description Markup letters ------- -----------
+ * -------------- SS Secondary Structure [HGIEBTSCX] SA Surface
+ * Accessibility [0-9X] (0=0%-10%; ...; 9=90%-100%) TM TransMembrane
+ * [Mio] PP Posterior Probability [0-9*] (0=0.00-0.05; 1=0.05-0.15;
+ * *=0.95-1.00) LI LIgand binding [*] AS Active Site [*] IN INtron (in
+ * or after) [0-2]
+ */
+ if (s.search(annContent))
+ {
+ String acc = s.stringMatched(1);
+ String type = s.stringMatched(2);
+ String oseq = s.stringMatched(3);
+ /*
+ * copy of annotation field that may be processed into whitespace chunks
+ */
+ String seq = new String(oseq);
+
+ Hashtable ann;
+ // Get an object with all the annotations for this sequence
+ if (seqAnn.containsKey(acc))
+ {
+ // logger.debug("Found annotations for " + acc);
+ ann = (Hashtable) seqAnn.get(acc);
+ }
+ else
+ {
+ // logger.debug("Creating new annotations holder for " + acc);
+ ann = new Hashtable();
+ seqAnn.put(acc, ann);
+ }
+
+ // // start of block for appending annotation lines for wrapped
+ // stokchholm file
+ // TODO test structure, call parseAnnotationRow with vector from
+ // hashtable for specific sequence
+
+ Hashtable features;
+ // Get an object with all the content for an annotation
+ if (ann.containsKey("features"))
+ {
+ // logger.debug("Found features for " + acc);
+ features = (Hashtable) ann.get("features");
+ }
+ else
+ {
+ // logger.debug("Creating new features holder for " + acc);
+ features = new Hashtable();
+ ann.put("features", features);
+ }
+
+ Hashtable content;
+ if (features.containsKey(this.id2type(type)))
+ {
+ // logger.debug("Found content for " + this.id2type(type));
+ content = (Hashtable) features.get(this.id2type(type));
+ }
+ else
+ {
+ // logger.debug("Creating new content holder for " +
+ // this.id2type(type));
+ content = new Hashtable();
+ features.put(this.id2type(type), content);
+ }
+ String ns = (String) content.get(ANNOTATION);
+
+ if (ns == null)
+ {
+ ns = "";
+ }
+ // finally, append the annotation line
+ ns += seq;
+ content.put(ANNOTATION, ns);
+ // // end of wrapped annotation block.
+ // // Now a new row is created with the current set of data
+
+ Hashtable strucAnn;
+ if (seqAnn.containsKey(acc))
+ {
+ strucAnn = (Hashtable) seqAnn.get(acc);
+ }
+ else
+ {
+ strucAnn = new Hashtable();
+ }
+
+ Vector<AlignmentAnnotation> newStruc = new Vector<>();
+ parseAnnotationRow(newStruc, type, ns);
+ for (AlignmentAnnotation alan : newStruc)
+ {
+ alan.visible = false;
+ }
+ // new annotation overwrites any existing annotation...
+
+ strucAnn.put(type, newStruc);
+ seqAnn.put(acc, strucAnn);
+ }
+ // }
+ else
+ {
+ jalview.bin.Console.errPrintln(
+ "Warning - couldn't parse sequence annotation row line:\n"
+ + line);
+ // throw new IOException("Error parsing " + line);
+ }
+ }
+ else
+ {
+ throw new IOException(MessageManager.formatMessage(
+ "exception.unknown_annotation_detected", new String[]
+ { annType, annContent }));
+ }
+ }
+ }
+ if (treeString.length() > 0)
+ {
+ if (treeName == null)
+ {
+ treeName = "Tree " + (1 + getTreeCount());
+ }
+ addNewickTree(treeName, treeString.toString());
+ }
+ }
+
+ /**
+ * Demangle an accession string and guess the originating sequence database
+ * for a given sequence
+ *
+ * @param seqO
+ * sequence to be annotated
+ * @param dbr
+ * Accession string for sequence
+ * @param dbsource
+ * source database for alignment (PFAM or RFAM)
+ */
+ private void guessDatabaseFor(Sequence seqO, String dbr, String dbsource)
+ {
+ DBRefEntry dbrf = null;
+ List<DBRefEntry> dbrs = new ArrayList<>();
+ String seqdb = "Unknown", sdbac = "" + dbr;
+ int st = -1, en = -1, p;
+ if ((st = sdbac.indexOf("/")) > -1)
+ {
+ String num, range = sdbac.substring(st + 1);
+ sdbac = sdbac.substring(0, st);
+ if ((p = range.indexOf("-")) > -1)
+ {
+ p++;
+ if (p < range.length())
+ {
+ num = range.substring(p).trim();
+ try
+ {
+ en = Integer.parseInt(num);
+ } catch (NumberFormatException x)
+ {
+ // could warn here that index is invalid
+ en = -1;
+ }
+ }
+ }
+ else
+ {
+ p = range.length();
+ }
+ num = range.substring(0, p).trim();
+ try
+ {
+ st = Integer.parseInt(num);
+ } catch (NumberFormatException x)
+ {
+ // could warn here that index is invalid
+ st = -1;
+ }
+ }
+ if (dbsource == null)
+ {
+ // make up an origin based on whether the sequence looks like it is
+ // nucleotide
+ // or protein
+ dbsource = (seqO.isProtein()) ? "PFAM" : "RFAM";
+ }
+ if (dbsource.equals("PFAM"))
+ {
+ seqdb = "UNIPROT";
+ if (sdbac.indexOf(".") > -1)
+ {
+ // strip of last subdomain
+ sdbac = sdbac.substring(0, sdbac.indexOf("."));
+ dbrf = jalview.util.DBRefUtils.parseToDbRef(seqO, seqdb, dbsource,
+ sdbac);
+ if (dbrf != null)
+ {
+ dbrs.add(dbrf);
+ }
+ }
+ dbrf = jalview.util.DBRefUtils.parseToDbRef(seqO, dbsource, dbsource,
+ dbr);
+ if (dbr != null)
+ {
+ dbrs.add(dbrf);
+ }
+ }
+ else
+ {
+ seqdb = "EMBL"; // total guess - could be ENA, or something else these
+ // days
+ if (sdbac.indexOf(".") > -1)
+ {
+ // strip off last subdomain
+ sdbac = sdbac.substring(0, sdbac.indexOf("."));
+ dbrf = jalview.util.DBRefUtils.parseToDbRef(seqO, seqdb, dbsource,
+ sdbac);
+ if (dbrf != null)
+ {
+ dbrs.add(dbrf);
+ }
+ }
+
+ dbrf = jalview.util.DBRefUtils.parseToDbRef(seqO, dbsource, dbsource,
+ dbr);
+ if (dbrf != null)
+ {
+ dbrs.add(dbrf);
+ }
+ }
+ if (st != -1 && en != -1)
+ {
+ for (DBRefEntry d : dbrs)
+ {
+ jalview.util.MapList mp = new jalview.util.MapList(
+ new int[]
+ { seqO.getStart(), seqO.getEnd() }, new int[] { st, en }, 1,
+ 1);
+ jalview.datamodel.Mapping mping = new Mapping(mp);
+ d.setMap(mping);
+ }
+ }
+ }
+
+ protected static AlignmentAnnotation parseAnnotationRow(
+ Vector<AlignmentAnnotation> annotation, String label,
+ String annots)
+ {
+ String convert1, convert2 = null;
+
+ // convert1 = OPEN_PAREN.replaceAll(annots);
+ // convert2 = CLOSE_PAREN.replaceAll(convert1);
+ // annots = convert2;
+