X-Git-Url: http://source.jalview.org/gitweb/?a=blobdiff_plain;f=src%2Fjalview%2Fio%2FStockholmFile.java;h=4697262a7143db6b8fb022a0c98fe26b910801e6;hb=121a4b153bdd56b15ff2c528e6b84219affc09d5;hp=f5b517728c85ce77ad785c9030e7b7e475ed2046;hpb=be762d8d9c71a7aa3121e845c45911c7192b7827;p=jalview.git diff --git a/src/jalview/io/StockholmFile.java b/src/jalview/io/StockholmFile.java index f5b5177..4697262 100644 --- a/src/jalview/io/StockholmFile.java +++ b/src/jalview/io/StockholmFile.java @@ -28,12 +28,14 @@ import jalview.datamodel.AlignmentAnnotation; import jalview.datamodel.AlignmentI; import jalview.datamodel.Annotation; import jalview.datamodel.DBRefEntry; +import jalview.datamodel.DBRefSource; import jalview.datamodel.Mapping; import jalview.datamodel.Sequence; import jalview.datamodel.SequenceFeature; import jalview.datamodel.SequenceI; import jalview.schemes.ResidueProperties; import jalview.util.Comparison; +import jalview.util.DBRefUtils; import jalview.util.Format; import jalview.util.MessageManager; @@ -83,6 +85,14 @@ public class StockholmFile extends AlignFile public static final Regex DETECT_BRACKETS = new Regex( "(<|>|\\[|\\]|\\(|\\)|\\{|\\})"); + // WUSS extended symbols. Avoid ambiguity with protein SS annotations by using NOT_RNASS first. + public static final String RNASS_BRACKETS = "<>[](){}AaBbCcDdEeFfGgHhIiJjKkLlMmNnOoPpQqRrSsTtUuVvWwXxYyZz"; + + // use the following regex to decide an annotations (whole) line is NOT an RNA + // SS (it contains only E,H,e,h and other non-brace/non-alpha chars) + private static final Regex NOT_RNASS = new Regex( + "^[^<>[\\](){}A-DF-Za-df-z]*$"); + StringBuffer out; // output buffer AlignmentI al; @@ -197,7 +207,7 @@ public class StockholmFile extends AlignFile String version; // String id; Hashtable seqAnn = new Hashtable(); // Sequence related annotations - LinkedHashMap seqs = new LinkedHashMap(); + LinkedHashMap seqs = new LinkedHashMap<>(); Regex p, r, rend, s, x; // Temporary line for processing RNA annotation // String RNAannot = ""; @@ -324,17 +334,14 @@ public class StockholmFile extends AlignFile if (accAnnotations != null && accAnnotations.containsKey("AC")) { - if (dbsource != null) + String dbr = (String) accAnnotations.get("AC"); + if (dbr != null) { - 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 source of alignment plus structure - // of accession - guessDatabaseFor(seqO, dbr, dbsource); - - } + // 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 ? @@ -519,6 +526,9 @@ public class StockholmFile extends AlignFile 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)); } } @@ -658,7 +668,7 @@ public class StockholmFile extends AlignFile strucAnn = new Hashtable(); } - Vector newStruc = new Vector(); + Vector newStruc = new Vector<>(); parseAnnotationRow(newStruc, type, ns); for (AlignmentAnnotation alan : newStruc) { @@ -710,7 +720,7 @@ public class StockholmFile extends AlignFile private void guessDatabaseFor(Sequence seqO, String dbr, String dbsource) { DBRefEntry dbrf = null; - List dbrs = new ArrayList(); + List dbrs = new ArrayList<>(); String seqdb = "Unknown", sdbac = "" + dbr; int st = -1, en = -1, p; if ((st = sdbac.indexOf("/")) > -1) @@ -747,6 +757,12 @@ public class StockholmFile extends AlignFile 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"; @@ -824,9 +840,14 @@ public class StockholmFile extends AlignFile } boolean ss = false, posterior = false; type = id2type(type); + + boolean isrnass = false; if (type.equalsIgnoreCase("secondary structure")) { ss = true; + isrnass = !NOT_RNASS.search(annots); // sorry about the double negative + // here (it's easier for dealing with + // other non-alpha-non-brace chars) } if (type.equalsIgnoreCase("posterior probability")) { @@ -844,7 +865,7 @@ public class StockholmFile extends AlignFile { // if (" .-_".indexOf(pos) == -1) { - if (DETECT_BRACKETS.search(pos)) + if (isrnass && RNASS_BRACKETS.indexOf(pos) >= 0) { ann.secondaryStructure = Rna.getRNASecStrucState(pos).charAt(0); ann.displayCharacter = "" + pos.charAt(0); @@ -917,6 +938,11 @@ public class StockholmFile extends AlignFile return annot; } + private String dbref_to_ac_record(DBRefEntry ref) + { + return ref.getSource().toString() + " ; " + + ref.getAccessionId().toString(); + } @Override public String print(SequenceI[] s, boolean jvSuffix) { @@ -929,8 +955,10 @@ public class StockholmFile extends AlignFile int maxid = 0; int in = 0; Hashtable dataRef = null; + boolean isAA = s[in].isProtein(); while ((in < s.length) && (s[in] != null)) { + String tmp = printId(s[in], jvSuffix); max = Math.max(max, s[in].getLength()); @@ -940,17 +968,33 @@ public class StockholmFile extends AlignFile } if (s[in].getDBRefs() != null) { - for (int idb = 0; idb < s[in].getDBRefs().length; idb++) + if (dataRef == null) + { + dataRef = new Hashtable(); + } + List primrefs = s[in].getPrimaryDBRefs(); + if (primrefs.size() >= 1) { - if (dataRef == null) + dataRef.put(tmp, dbref_to_ac_record(primrefs.get(0))); + } + else + { + for (int idb = 0; idb < s[in].getDBRefs().length; idb++) { - dataRef = new Hashtable(); + DBRefEntry dbref = s[in].getDBRefs()[idb]; + dataRef.put(tmp, dbref_to_ac_record(dbref)); + // if we put in a uniprot or EMBL record then we're done: + if (isAA && DBRefSource.UNIPROT + .equals(DBRefUtils.getCanonicalName(dbref.getSource()))) + { + break; + } + if (!isAA && DBRefSource.EMBL + .equals(DBRefUtils.getCanonicalName(dbref.getSource()))) + { + break; + } } - - String datAs1 = s[in].getDBRefs()[idb].getSource().toString() - + " ; " - + s[in].getDBRefs()[idb].getAccessionId().toString(); - dataRef.put(tmp, datAs1); } } in++; @@ -983,7 +1027,8 @@ public class StockholmFile extends AlignFile String type = (String) dataRef.remove(idd); out.append(new Format("%-" + (maxid - 2) + "s") .form("#=GS " + idd.toString() + " ")); - if (type.contains("PFAM") || type.contains("RFAM")) + if (isAA && type.contains("UNIPROT") + || (!isAA && type.contains("EMBL"))) { out.append(" AC " + type.substring(type.indexOf(";") + 1)); @@ -1114,22 +1159,36 @@ public class StockholmFile extends AlignFile String ch = (annot == null) ? ((sequenceI == null) ? "-" : Character.toString(sequenceI.getCharAt(k))) - : annot.displayCharacter; + : (annot.displayCharacter == null + ? String.valueOf(annot.secondaryStructure) + : annot.displayCharacter); + if (ch == null) + { + ch = " "; + } if (key != null && key.equals("SS")) { + char ssannotchar = ' '; + boolean charset = false; if (annot == null) { // sensible gap character - return ' '; + ssannotchar = ' '; + charset = true; } else { // valid secondary structure AND no alternative label (e.g. ' B') if (annot.secondaryStructure > ' ' && ch.length() < 2) { - return annot.secondaryStructure; + ssannotchar = annot.secondaryStructure; + charset = true; } } + if (charset) + { + return (ssannotchar == ' ' && isrna) ? '.' : ssannotchar; + } } if (ch.length() == 0) @@ -1144,7 +1203,9 @@ public class StockholmFile extends AlignFile { seq = ch.charAt(1); } - return seq; + + return (seq == ' ' && key != null && key.equals("SS") && isrna) ? '.' + : seq; } public String print()