import jalview.datamodel.AlignmentAnnotation;\r
import jalview.datamodel.Annotation;\r
import jalview.datamodel.Sequence;\r
+import jalview.datamodel.SequenceFeature;\r
import jalview.datamodel.SequenceI;\r
\r
- import java.io.*;\r
- import java.util.*;\r
- \r
- \r
- import javax.xml.parsers.ParserConfigurationException;\r
- \r
- import org.xml.sax.SAXException;\r
+ import java.io.BufferedReader;\r
+ import java.io.FileReader;\r
+ import java.io.IOException;\r
+ import java.util.ArrayList;\r
+ import java.util.Enumeration;\r
+ import java.util.Hashtable;\r
++import java.util.StringTokenizer;\r
+ import java.util.Vector;\r
\r
import com.stevesoft.pat.Regex;\r
\r
*/\r
public void parse() throws IOException\r
{\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
+ // 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
+ \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
+ \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
+ \r
+ this.noSeqs = seqs.size();\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\r
+ * Associate 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
+ // 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
+ // seqO.addDBRef(dbref);\r
+ }\r
+ } \r
+ if (accAnnotations != null && accAnnotations.containsKey("SS"))\r
+ {\r
+ Vector v = (Vector) accAnnotations.get("SS");\r
+ \r
+ for (int i = 0; i < v.size(); i++)\r
+ {\r
+ AlignmentAnnotation an = (AlignmentAnnotation) v.elementAt(i);\r
+ seqO.addAlignmentAnnotation(an);\r
+ //annotations.add(an);\r
+ }\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
+ 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 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
+
+ if(type.equals("SS")){\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
+ \r
+ strucAnn.put(type, newStruc);\r
+ seqAnn.put(acc, strucAnn);\r
+ }\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
- protected static AlignmentAnnotation parseAnnotationRow(Vector annotation,\r
- String label, String annots)\r
++\r
++ /**\r
++ * Parse a file in Stockholm format into Jalview's data model using VARNA\r
++ * \r
++ * @throws IOException\r
++ * If there is an error with the input file\r
++ */\r
++ public void parse_with_VARNA() 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
+ protected static AlignmentAnnotation parseAnnotationRow(\r
+ Vector annotation, String label, String annots)\r
{\r
String convert1, convert2 = null;\r
\r
private String safeName(String dataName)\r
{\r
int b = 0;\r
- while ((b = dataName.indexOf("/")) > -1 && b < dataName.length() )\r
- { \r
- dataName = dataName.substring(b + 1).trim();\r
- \r
+ while ((b = dataName.indexOf("/")) > -1 && b < dataName.length())\r
+ {\r
+ dataName = dataName.substring(b + 1).trim();\r
+ \r
}\r
- int e = (dataName.length() - dataName.indexOf("."))+1;\r
- dataName = dataName.substring(1,e).trim();\r
+ int e = (dataName.length() - dataName.indexOf(".")) + 1;\r
+ dataName = dataName.substring(1, e).trim();\r
return dataName;\r
}\r
-}
+}\r
git://source.jalview.org / jalview.git / commitdiff