import jalview.datamodel.SequenceI;
import jalview.schemes.ResidueProperties;
import jalview.util.Comparison;
+import jalview.util.DBRefUtils;
import jalview.util.Format;
import jalview.util.MessageManager;
+import jalview.util.Platform;
import java.io.BufferedReader;
import java.io.FileReader;
import java.io.IOException;
import java.util.ArrayList;
import java.util.Enumeration;
-import java.util.HashMap;
import java.util.Hashtable;
import java.util.LinkedHashMap;
import java.util.List;
import java.util.Map;
-import java.util.Map.Entry;
import java.util.Vector;
import com.stevesoft.pat.Regex;
private static final String ANNOTATION = "annotation";
private static final char UNDERSCORE = '_';
+
+ // WUSS extended symbols. Avoid ambiguity with protein SS annotations by using NOT_RNASS first.
- // private static final Regex OPEN_PAREN = new Regex("(<|\\[)", "(");
- // private static final Regex CLOSE_PAREN = new Regex("(>|\\])", ")");
+ public static final String RNASS_BRACKETS = "<>[](){}AaBbCcDdEeFfGgHhIiJjKkLlMmNnOoPpQqRrSsTtUuVvWwXxYyZz";
- public static final Regex DETECT_BRACKETS = new Regex(
- "(<|>|\\[|\\]|\\(|\\)|\\{|\\})");
+ public static final int REGEX_STOCKHOLM = 0;
- /*
- * lookup table of Stockholm 'feature' (annotation) types
- * see http://sonnhammer.sbc.su.se/Stockholm.html
- */
- private static Map<String, String> featureTypes = null;
-
- // WUSS extended symbols. Avoid ambiguity with protein SS annotations by using NOT_RNASS first.
- public static final String RNASS_BRACKETS = "<>[](){}AaBbCcDdEeFfGgHhIiJjKkLlMmNnOoPpQqRrSsTtUuVvWwXxYyZz";
+ public static final int REGEX_BRACKETS = 1;
// 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]*$");
+ public static final int REGEX_NOT_RNASS = 2;
- StringBuffer out; // output buffer
+ private static final int REGEX_ANNOTATION = 3;
- static
+ private static final int REGEX_PFAM = 4;
+
+ private static final int REGEX_RFAM = 5;
+
+ private static final int REGEX_ALIGN_END = 6;
+
+ private static final int REGEX_SPLIT_ID = 7;
+
+ private static final int REGEX_SUBTYPE = 8;
+
+ private static final int REGEX_ANNOTATION_LINE = 9;
+
+ private static final int REGEX_REMOVE_ID = 10;
+
+ private static final int REGEX_OPEN_PAREN = 11;
+
+ private static final int REGEX_CLOSE_PAREN = 12;
+
+ public static final int REGEX_MAX = 13;
+
+ private static Regex REGEX[] = new Regex[REGEX_MAX];
+
+ /**
+ * Centralize all actual Regex instantialization in Platform.
+ * // JBPNote: Why is this 'centralisation' better ?
+ * @param id
+ * @return
+ */
+ private static Regex getRegex(int id)
{
- featureTypes = new HashMap<>();
- featureTypes.put("SS", "Secondary Structure");
- featureTypes.put("SA", "Surface Accessibility");
- featureTypes.put("TM", "transmembrane");
- featureTypes.put("PP", "Posterior Probability");
- featureTypes.put("LI", "ligand binding");
- featureTypes.put("AS", "active site");
- featureTypes.put("IN", "intron");
- featureTypes.put("IR", "interacting residue");
- featureTypes.put("AC", "accession");
- featureTypes.put("OS", "organism");
- featureTypes.put("CL", "class");
- featureTypes.put("DE", "description");
- featureTypes.put("DR", "reference");
- featureTypes.put("LO", "look");
- featureTypes.put("RF", "Reference Positions");
+ if (REGEX[id] == null)
+ {
+ String pat = null, pat2 = null;
+ switch (id)
+ {
+ case REGEX_STOCKHOLM:
+ pat = "# STOCKHOLM ([\\d\\.]+)";
+ break;
+ case REGEX_BRACKETS:
+ // for reference; not used
+ pat = "(<|>|\\[|\\]|\\(|\\)|\\{|\\})";
+ break;
+ case REGEX_NOT_RNASS:
+ pat = "^[^<>[\\](){}A-DF-Za-df-z]*$";
+ break;
+ case REGEX_ANNOTATION:
+ pat = "(\\w+)\\s*(.*)";
+ break;
+ case REGEX_PFAM:
+ pat = "PF[0-9]{5}(.*)";
+ break;
+ case REGEX_RFAM:
+ pat = "RF[0-9]{5}(.*)";
+ break;
+ case REGEX_ALIGN_END:
+ pat = "^\\s*\\/\\/";
+ break;
+ case REGEX_SPLIT_ID:
+ pat = "(\\S+)\\/(\\d+)\\-(\\d+)";
+ break;
+ case REGEX_SUBTYPE:
+ pat = "(\\S+)\\s+(\\S*)\\s+(.*)";
+ break;
+ case REGEX_ANNOTATION_LINE:
+ pat = "#=(G[FSRC]?)\\s+(.*)";
+ break;
+ case REGEX_REMOVE_ID:
+ pat = "(\\S+)\\s+(\\S+)";
+ break;
+ case REGEX_OPEN_PAREN:
+ pat = "(<|\\[)";
+ pat2 = "(";
+ break;
+ case REGEX_CLOSE_PAREN:
+ pat = "(>|\\])";
+ pat2 = ")";
+ break;
+ default:
+ return null;
+ }
+ REGEX[id] = Platform.newRegex(pat, pat2);
+ }
+ return REGEX[id];
}
+ StringBuffer out; // output buffer
+
private AlignmentI al;
public StockholmFile()
super(source);
}
- /**
- * Answers the readable description for a (case-sensitive) annotation type
- * code, for example "Reference Positions" for "RF". Returns the type code if
- * no description is found.
- *
- * @param id
- * @return
- */
- public static String typeToDescription(String id)
- {
- if (featureTypes.containsKey(id))
- {
- return featureTypes.get(id);
- }
- System.err.println(
- "Warning : Unknown Stockholm annotation type code " + id);
- return id;
- }
-
- /**
- * Answers the annotation type code for a (non-case-sensitive) readable
- * description, for example "RF" for "Reference Positions" (or null if not
- * found)
- *
- * @param description
- * @return
- */
- public static String descriptionToType(String description)
- {
- for (Entry<String, String> entry : featureTypes.entrySet())
- {
- if (entry.getValue().equalsIgnoreCase(description))
- {
- return entry.getKey();
- }
- }
- System.err.println(
- "Warning : Unknown Stockholm annotation type: " + description);
- return null;
- }
-
@Override
public void initData()
{
// First, we have to check that this file has STOCKHOLM format, i.e. the
// first line must match
- r = new Regex("# STOCKHOLM ([\\d\\.]+)");
+ r = getRegex(REGEX_STOCKHOLM);
if (!r.search(nextLine()))
{
throw new IOException(MessageManager
// logger.debug("Stockholm version: " + version);
}
- // We define some Regexes here that will be used regularly later
- rend = new Regex("^\\s*\\/\\/"); // Find the end of an alignment
- p = new Regex("(\\S+)\\/(\\d+)\\-(\\d+)"); // split sequence id in
+ // We define some Regexes here that will be used regularily later
+ rend = getRegex(REGEX_ALIGN_END);//"^\\s*\\/\\/"); // Find the end of an alignment
+ p = getRegex(REGEX_SPLIT_ID);//"(\\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
+ s = getRegex(REGEX_SUBTYPE);// "(\\S+)\\s+(\\S*)\\s+(.*)"); // Parses
+ // annotation subtype
+ r = getRegex(REGEX_ANNOTATION_LINE);// "#=(G[FSRC]?)\\s+(.*)"); // Finds any
+ // annotation line
+ x = getRegex(REGEX_REMOVE_ID);// "(\\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("(>|\\])", ")");
+ Regex openparen = getRegex(REGEX_OPEN_PAREN);//"(<|\\[)", "(");
+ Regex closeparen = getRegex(REGEX_CLOSE_PAREN);//"(>|\\])", ")");
- // Detect if file is RNA by looking for bracket types
- // Regex detectbrackets = new Regex("(<|>|\\[|\\]|\\(|\\))");
+// // Detect if file is RNA by looking for bracket types
+ // Regex detectbrackets = getRegex("(<|>|\\[|\\]|\\(|\\))");
rend.optimize();
p.optimize();
// End of the alignment, pass stuff back
this.noSeqs = seqs.size();
- String seqdb, 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
+ String dbsource = null;
+ Regex pf = getRegex(REGEX_PFAM); // Finds AC for Pfam
+ Regex rf = getRegex(REGEX_RFAM); // Finds AC for Rfam
if (getAlignmentProperty("AC") != null)
{
String dbType = getAlignmentProperty("AC").toString();
if (accAnnotations != null && accAnnotations.containsKey("AC"))
{
- if (dbsource != 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
+ // 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 content = (Hashtable) features.remove(type);
// add alignment annotation for this feature
- String key = descriptionToType(type);
+ String key = type2id(type);
/*
* have we added annotation rows for this type ?
*/
// Let's save the annotations, maybe we'll be able to do something
// with them later...
- Regex an = new Regex("(\\w+)\\s*(.*)");
+ Regex an = getRegex(REGEX_ANNOTATION);
if (an.search(annContent))
{
if (an.stringMatched(1).equals("NH"))
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));
}
}
}
Hashtable content;
- if (features.containsKey(StockholmFile.typeToDescription(type)))
+ if (features.containsKey(this.id2type(type)))
{
// logger.debug("Found content for " + this.id2type(type));
content = (Hashtable) features
- .get(StockholmFile.typeToDescription(type));
+ .get(this.id2type(type));
}
else
{
// logger.debug("Creating new content holder for " +
// this.id2type(type));
content = new Hashtable();
- features.put(StockholmFile.typeToDescription(type), content);
+ features.put(id2type(type), content);
}
String ns = (String) content.get(ANNOTATION);
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";
Vector<AlignmentAnnotation> annotation, String label,
String annots)
{
+ String convert1, convert2 = null;
// String convert1 = OPEN_PAREN.replaceAll(annots);
// String convert2 = CLOSE_PAREN.replaceAll(convert1);
// annots = convert2;
if (type.equalsIgnoreCase("secondary structure"))
{
ss = true;
- isrnass = !NOT_RNASS.search(annots); // sorry about the double negative
+ isrnass = !getRegex(REGEX_NOT_RNASS).search(annots); // sorry about the double
+ // negative
// here (it's easier for dealing with
// other non-alpha-non-brace chars)
}
return annot;
}
+ private String dbref_to_ac_record(DBRefEntry ref)
+ {
+ return ref.getSource().toString() + " ; "
+ + ref.getAccessionId().toString();
+ }
+
@Override
- public String print(final SequenceI[] sequences, boolean jvSuffix)
+ public String print(SequenceI[] s, boolean jvSuffix)
{
- StringBuilder out = new StringBuilder();
+ out = new StringBuffer();
out.append("# STOCKHOLM 1.0");
out.append(newline);
- int maxIdWidth = 0;
- for (SequenceI seq : sequences)
+ // find max length of id
+ int max = 0;
+ int maxid = 0;
+ int in = 0;
+ int slen = s.length;
+ SequenceI seq;
+ Hashtable<String, String> dataRef = null;
+ boolean isAA = s[in].isProtein();
+ while ((in < slen) && ((seq = s[in]) != null))
{
- if (seq != null)
+ String tmp = printId(seq, jvSuffix);
+ max = Math.max(max, seq.getLength());
+
+ if (tmp.length() > maxid)
+ {
+ maxid = tmp.length();
+ }
+ List<DBRefEntry> seqrefs = seq.getDBRefs();
+ int ndb;
+ if (seqrefs != null && (ndb = seqrefs.size()) > 0)
{
- String formattedId = printId(seq, jvSuffix);
- maxIdWidth = Math.max(maxIdWidth, formattedId.length());
+ if (dataRef == null)
+ {
+ dataRef = new Hashtable<>();
+ }
+ List<DBRefEntry> primrefs = seq.getPrimaryDBRefs();
+ if (primrefs.size() >= 1)
+ {
+ dataRef.put(tmp, dbref_to_ac_record(primrefs.get(0)));
+ }
+ else
+ {
+ for (int idb = 0; idb < ndb; idb++)
+ {
+ DBRefEntry dbref = seqrefs.get(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 : DBRefSource.EMBL)
+ .equals(DBRefUtils.getCanonicalName(dbref.getSource())))
+ {
+ break;
+ }
+ }
+ }
}
+ in++;
}
- maxIdWidth += 9;
+ maxid += 9;
+ int i = 0;
- /*
- * generic alignment properties
- */
- Hashtable props = al.getProperties();
- if (props != null)
+ // output database type
+ if (al.getProperties() != null)
{
- for (Object key : props.keySet())
+ if (!al.getProperties().isEmpty())
{
- out.append(String.format("#=GF %s %s", key.toString(),
- props.get(key).toString()));
- out.append(newline);
+ Enumeration key = al.getProperties().keys();
+ Enumeration val = al.getProperties().elements();
+ while (key.hasMoreElements())
+ {
+ out.append("#=GF " + key.nextElement() + " " + val.nextElement());
+ out.append(newline);
+ }
}
}
- /*
- * output database accessions as #=GS (per sequence annotation)
- * PFAM or RFAM are output as AC <accession number>
- * others are output as DR <dbname> ; <accession>
- */
- Format formatter = new Format("%-" + (maxIdWidth - 2) + "s");
- for (SequenceI seq : sequences)
+ // output database accessions
+ if (dataRef != null)
{
- if (seq != null)
+ Enumeration<String> en = dataRef.keys();
+ while (en.hasMoreElements())
{
- DBRefEntry[] dbRefs = seq.getDBRefs();
- if (dbRefs != null)
+ Object idd = en.nextElement();
+ String type = dataRef.remove(idd);
+ out.append(new Format("%-" + (maxid - 2) + "s")
+ .form("#=GS " + idd.toString() + " "));
+ if (isAA && type.contains("UNIPROT")
+ || (!isAA && type.contains("EMBL")))
{
- String idField = formatter
- .form("#=GS " + printId(seq, jvSuffix) + " ");
- for (DBRefEntry dbRef : dbRefs)
- {
- out.append(idField);
- printDbRef(out, dbRef);
- }
+
+ out.append(" AC " + type.substring(type.indexOf(";") + 1));
}
+ else
+ {
+ out.append(" DR " + type + " ");
+ }
+ out.append(newline);
}
}
- /*
- * output annotations
- */
- for (SequenceI seq : sequences)
+ // output annotations
+ while (i < slen && (seq = s[i]) != null)
{
- if (seq != null)
+ AlignmentAnnotation[] alAnot = seq.getAnnotation();
+ if (alAnot != null)
{
- AlignmentAnnotation[] alAnot = seq.getAnnotation();
- if (alAnot != null)
+ Annotation[] ann;
+ for (int j = 0; j < alAnot.length; j++)
{
- for (int j = 0; j < alAnot.length; j++)
+ if (alAnot[j].annotations != null)
{
- AlignmentAnnotation ann = alAnot[j];
- String key = descriptionToType(ann.label);
- boolean isrna = ann.isValidStruc();
+ String key = type2id(alAnot[j].label);
+ boolean isrna = alAnot[j].isValidStruc();
+
if (isrna)
{
- /*
- * output as secondary structure if there is
- * RNA secondary structure on the annotation
- */
+ // hardwire to secondary structure if there is RNA secondary
+ // structure on the annotation
key = "SS";
}
if (key == null)
continue;
}
- out.append(new Format("%-" + maxIdWidth + "s").form(
- "#=GR " + printId(seq, jvSuffix) + " " + key + " "));
- Annotation[] anns = ann.annotations;
- StringBuilder seqString = new StringBuilder();
- for (int k = 0; k < anns.length; k++)
+ // out.append("#=GR ");
+ out.append(new Format("%-" + maxid + "s").form(
+ "#=GR " + printId(s[i], jvSuffix) + " " + key + " "));
+ ann = alAnot[j].annotations;
+ String sseq = "";
+ for (int k = 0; k < ann.length; k++)
{
- seqString
- .append(getAnnotationCharacter(key, k, anns[k], seq));
+ sseq += outputCharacter(key, k, isrna, ann, s[i]);
}
- out.append(seqString.toString());
+ out.append(sseq);
out.append(newline);
- }
+ }
}
- out.append(new Format("%-" + maxIdWidth + "s")
- .form(printId(seq, jvSuffix) + " "));
- out.append(seq.getSequenceAsString());
- out.append(newline);
}
+
+ out.append(new Format("%-" + maxid + "s")
+ .form(printId(seq, jvSuffix) + " "));
+ out.append(seq.getSequenceAsString());
+ out.append(newline);
+ i++;
}
- /*
- * output alignment annotation (but not auto-calculated or sequence-related)
- */
- if (al.getAlignmentAnnotation() != null)
+ // alignment annotation
+ AlignmentAnnotation aa;
+ AlignmentAnnotation[] an = al.getAlignmentAnnotation();
+ if (an != null)
{
- for (int ia = 0; ia < al.getAlignmentAnnotation().length; ia++)
+ for (int ia = 0, na = an.length; ia < na; ia++)
{
- AlignmentAnnotation aa = al.getAlignmentAnnotation()[ia];
+ aa = an[ia];
if (aa.autoCalculated || !aa.visible || aa.sequenceRef != null)
{
continue;
}
- String label = aa.label;
+ String sseq = "";
+ String label;
String key = "";
if (aa.label.equals("seq"))
{
}
else
{
- key = descriptionToType(aa.label);
- if ("RF".equals(key))
+ key = type2id(aa.label.toLowerCase());
+ if (key == null)
{
- label = key;
+ label = aa.label;
}
- else if (key != null)
+ else
{
label = key + "_cons";
}
}
+ if (label == null)
+ {
+ label = aa.label;
+ }
label = label.replace(" ", "_");
out.append(
- new Format("%-" + maxIdWidth + "s")
- .form("#=GC " + label + " "));
- StringBuilder sb = new StringBuilder(aa.annotations.length);
- for (int j = 0; j < aa.annotations.length; j++)
+ new Format("%-" + maxid + "s").form("#=GC " + label + " "));
+ boolean isrna = aa.isValidStruc();
+ for (int j = 0, nj = aa.annotations.length; j < nj; j++)
{
- sb.append(
- getAnnotationCharacter(key, j, aa.annotations[j], null));
+ sseq += outputCharacter(key, j, isrna, aa.annotations, null);
}
- out.append(sb.toString());
+ out.append(sseq);
out.append(newline);
}
}
return out.toString();
}
- /**
- * A helper method that appends a formatted dbref to the output buffer
- *
- * @param out
- * @param dbRef
- */
- protected void printDbRef(StringBuilder out, DBRefEntry dbRef)
- {
- String db = dbRef.getSource();
- String acc = dbRef.getAccessionId();
- if (DBRefSource.PFAM.equalsIgnoreCase(db)
- || DBRefSource.RFAM.equalsIgnoreCase(db))
- {
- out.append(" AC " + acc);
- }
- else
- {
- out.append(" DR " + db + " ; " + acc);
- }
- out.append(newline);
- }
/**
- * Returns an annotation character to add to the output row
+ * add an annotation character to the output row
*
* @param seq
* @param key
* @param k
+ * @param isrna
* @param ann
* @param sequenceI
*/
- static char getAnnotationCharacter(String key, int k, Annotation annot,
- SequenceI sequenceI)
+ private char outputCharacter(String key, int k, boolean isrna,
+ Annotation[] ann, SequenceI sequenceI)
{
char seq = ' ';
+ Annotation annot = ann[k];
String ch = (annot == null)
? ((sequenceI == null) ? "-"
: Character.toString(sequenceI.getCharAt(k)))
boolean charset = false;
if (annot == null)
{
- // TODO: TVA: Check against develop for correct behaviour!
- // Stockholm format requires underscore, not space
- return UNDERSCORE;
+ // sensible gap character
+ ssannotchar = ' ';
+ charset = true;
}
else
{
dataName = dataName.substring(1, e).trim();
return dataName;
}
+
+
+ public String print()
+ {
+ out = new StringBuffer();
+ out.append("# STOCKHOLM 1.0");
+ out.append(newline);
+ print(getSeqsAsArray(), false);
+
+ out.append("//");
+ out.append(newline);
+ return out.toString();
+ }
+
+ private static Hashtable typeIds = null;
+
+ static
+ {
+ if (typeIds == null)
+ {
+ typeIds = new Hashtable();
+ typeIds.put("SS", "Secondary Structure");
+ typeIds.put("SA", "Surface Accessibility");
+ typeIds.put("TM", "transmembrane");
+ typeIds.put("PP", "Posterior Probability");
+ typeIds.put("LI", "ligand binding");
+ typeIds.put("AS", "active site");
+ typeIds.put("IN", "intron");
+ typeIds.put("IR", "interacting residue");
+ typeIds.put("AC", "accession");
+ typeIds.put("OS", "organism");
+ typeIds.put("CL", "class");
+ typeIds.put("DE", "description");
+ typeIds.put("DR", "reference");
+ typeIds.put("LO", "look");
+ typeIds.put("RF", "Reference Positions");
+
+ }
+ }
+
+ protected static String id2type(String id)
+ {
+ if (typeIds.containsKey(id))
+ {
+ return (String) typeIds.get(id);
+ }
+ System.err.println(
+ "Warning : Unknown Stockholm annotation type code " + id);
+ return id;
+ }
+
+ protected static String type2id(String type)
+ {
+ String key = null;
+ Enumeration e = typeIds.keys();
+ while (e.hasMoreElements())
+ {
+ Object ll = e.nextElement();
+ if (typeIds.get(ll).toString().equalsIgnoreCase(type))
+ {
+ key = (String) ll;
+ break;
+ }
+ }
+ if (key != null)
+ {
+ return key;
+ }
+ System.err.println(
+ "Warning : Unknown Stockholm annotation type: " + type);
+ return key;
+ }
}
git://source.jalview.org / jalview.git / blobdiff