X-Git-Url: http://source.jalview.org/gitweb/?a=blobdiff_plain;f=src%2Fjalview%2Fanalysis%2FAlignSeq.java;h=65fd1100b8e8a7f2cdc33f1d1a90f80aaa3076bf;hb=57738a1f3c19b1c3a00bd3ac5108f8cd0af32f99;hp=1a6826d7754a33cb0168190bc087eda935b50394;hpb=b0f76adef2787dd14566525e66a4073278e75d67;p=jalview.git diff --git a/src/jalview/analysis/AlignSeq.java b/src/jalview/analysis/AlignSeq.java index 1a6826d..65fd110 100755 --- a/src/jalview/analysis/AlignSeq.java +++ b/src/jalview/analysis/AlignSeq.java @@ -20,6 +20,8 @@ */ package jalview.analysis; +import java.util.Locale; + import jalview.analysis.scoremodels.PIDModel; import jalview.analysis.scoremodels.ScoreMatrix; import jalview.analysis.scoremodels.ScoreModels; @@ -36,6 +38,7 @@ import jalview.util.MessageManager; import java.awt.Color; import java.awt.Graphics; +import java.io.PrintStream; import java.util.ArrayList; import java.util.Arrays; import java.util.List; @@ -49,6 +52,14 @@ import java.util.StringTokenizer; */ public class AlignSeq { + private static final int MAX_NAME_LENGTH = 30; + + private static final int GAP_OPEN_COST = 120; + + private static final int GAP_EXTEND_COST = 20; + + private static final int GAP_INDEX = -1; + public static final String PEP = "pep"; public static final String DNA = "dna"; @@ -61,7 +72,7 @@ public class AlignSeq float[][] F; - int[][] traceback; + int[][] traceback; // todo is this actually used? int[] seq1; @@ -96,38 +107,29 @@ public class AlignSeq /** DOCUMENT ME!! */ public int seq2start; - /** DOCUMENT ME!! */ public int seq2end; int count; - /** DOCUMENT ME!! */ public float maxscore; - float pid; - int prev = 0; - int gapOpen = 120; - - int gapExtend = 20; - - float[][] lookup; - - int gapIndex = 23; - StringBuffer output = new StringBuffer(); String type; // AlignSeq.PEP or AlignSeq.DNA - private ScoreMatrix scoreModel; + private ScoreMatrix scoreMatrix; /** * Creates a new AlignSeq object. * - * @param s1 first sequence for alignment - * @param s2 second sequence for alignment - * @param type molecule type, either AlignSeq.PEP or AlignSeq.DNA + * @param s1 + * first sequence for alignment + * @param s2 + * second sequence for alignment + * @param type + * molecule type, either AlignSeq.PEP or AlignSeq.DNA */ public AlignSeq(SequenceI s1, SequenceI s2, String type) { @@ -148,7 +150,8 @@ public class AlignSeq public AlignSeq(SequenceI s1, String string1, SequenceI s2, String string2, String type) { - seqInit(s1, string1.toUpperCase(), s2, string2.toUpperCase(), type); + seqInit(s1, string1.toUpperCase(Locale.ROOT), s2, + string2.toUpperCase(Locale.ROOT), type); } /** @@ -260,8 +263,8 @@ public class AlignSeq SequenceI alSeq1 = new Sequence(s1.getName(), getAStr1()); alSeq1.setStart(s1.getStart() + getSeq1Start() - 1); alSeq1.setEnd(s1.getStart() + getSeq1End() - 1); - alSeq1.setDatasetSequence(s1.getDatasetSequence() == null ? s1 : s1 - .getDatasetSequence()); + alSeq1.setDatasetSequence( + s1.getDatasetSequence() == null ? s1 : s1.getDatasetSequence()); return alSeq1; } @@ -274,8 +277,8 @@ public class AlignSeq SequenceI alSeq2 = new Sequence(s2.getName(), getAStr2()); alSeq2.setStart(s2.getStart() + getSeq2Start() - 1); alSeq2.setEnd(s2.getStart() + getSeq2End() - 1); - alSeq2.setDatasetSequence(s2.getDatasetSequence() == null ? s2 : s2 - .getDatasetSequence()); + alSeq2.setDatasetSequence( + s2.getDatasetSequence() == null ? s2 : s2.getDatasetSequence()); return alSeq2; } @@ -316,16 +319,12 @@ public class AlignSeq if (s1str.length() == 0 || s2str.length() == 0) { - output.append("ALL GAPS: " - + (s1str.length() == 0 ? s1.getName() : " ") - + (s2str.length() == 0 ? s2.getName() : "")); + output.append( + "ALL GAPS: " + (s1str.length() == 0 ? s1.getName() : " ") + + (s2str.length() == 0 ? s2.getName() : "")); return; } - seq1 = new int[s1str.length()]; - - seq2 = new int[s2str.length()]; - score = new float[s1str.length()][s2str.length()]; E = new float[s1str.length()][s2str.length()]; @@ -343,16 +342,14 @@ public class AlignSeq if (!PEP.equals(moleculeType) && !DNA.equals(moleculeType)) { output.append("Wrong type = dna or pep only"); - throw new Error(MessageManager.formatMessage( - "error.unknown_type_dna_or_pep", - new String[] { moleculeType })); + throw new Error(MessageManager + .formatMessage("error.unknown_type_dna_or_pep", new String[] + { moleculeType })); } type = moleculeType; - scoreModel = ScoreModels.getInstance().getDefaultModel( - PEP.equals(type)); - lookup = scoreModel.getMatrix(); - gapIndex = scoreModel.getMatrixIndex(' '); + scoreMatrix = ScoreModels.getInstance() + .getDefaultModel(PEP.equals(type)); } /** @@ -383,11 +380,10 @@ public class AlignSeq } } - // System.out.println(maxi + " " + maxj + " " + score[maxi][maxj]); int i = maxi; int j = maxj; int trace; - maxscore = score[i][j] / 10; + maxscore = score[i][j] / 10f; seq1end = maxi + 1; seq2end = maxj + 1; @@ -417,13 +413,13 @@ public class AlignSeq else if (trace == 1) { j--; - aseq1[count] = gapIndex; + aseq1[count] = GAP_INDEX; sb1.replace(sb1.length() - 1, sb1.length(), "-"); } else if (trace == -1) { i--; - aseq2[count] = gapIndex; + aseq2[count] = GAP_INDEX; sb2.replace(sb2.length() - 1, sb2.length(), "-"); } @@ -433,13 +429,13 @@ public class AlignSeq seq1start = i + 1; seq2start = j + 1; - if (aseq1[count] != gapIndex) + if (aseq1[count] != GAP_INDEX) { aseq1[count] = seq1[i]; sb1.append(s1str.charAt(i)); } - if (aseq2[count] != gapIndex) + if (aseq2[count] != GAP_INDEX) { aseq2[count] = seq2[j]; sb2.append(s2str.charAt(j)); @@ -456,49 +452,48 @@ public class AlignSeq /** * DOCUMENT ME! */ - public void printAlignment(java.io.PrintStream os) + public void printAlignment(PrintStream os) { // TODO: Use original sequence characters rather than re-translated // characters in output // Find the biggest id length for formatting purposes - String s1id = s1.getName(), s2id = s2.getName(); - int maxid = s1.getName().length(); - if (s2.getName().length() > maxid) + String s1id = getAlignedSeq1().getDisplayId(true); + String s2id = getAlignedSeq2().getDisplayId(true); + int nameLength = Math.max(s1id.length(), s2id.length()); + if (nameLength > MAX_NAME_LENGTH) { - maxid = s2.getName().length(); - } - if (maxid > 30) - { - maxid = 30; + int truncateBy = nameLength - MAX_NAME_LENGTH; + nameLength = MAX_NAME_LENGTH; // JAL-527 - truncate the sequence ids - if (s1.getName().length() > maxid) + if (s1id.length() > nameLength) { - s1id = s1.getName().substring(0, 30); + int slashPos = s1id.lastIndexOf('/'); + s1id = s1id.substring(0, slashPos - truncateBy) + + s1id.substring(slashPos); } - if (s2.getName().length() > maxid) + if (s2id.length() > nameLength) { - s2id = s2.getName().substring(0, 30); + int slashPos = s2id.lastIndexOf('/'); + s2id = s2id.substring(0, slashPos - truncateBy) + + s2id.substring(slashPos); } } - int len = 72 - maxid - 1; + int len = 72 - nameLength - 1; int nochunks = ((aseq1.length - count) / len) + ((aseq1.length - count) % len > 0 ? 1 : 0); - pid = 0; + float pid = 0f; output.append("Score = ").append(score[maxi][maxj]).append(NEWLINE); output.append("Length of alignment = ") .append(String.valueOf(aseq1.length - count)).append(NEWLINE); output.append("Sequence "); - output.append(new Format("%" + maxid + "s").form(s1.getName())); - output.append(" : ").append(String.valueOf(s1.getStart())) - .append(" - ").append(String.valueOf(s1.getEnd())); + Format nameFormat = new Format("%" + nameLength + "s"); + output.append(nameFormat.form(s1id)); output.append(" (Sequence length = ") .append(String.valueOf(s1str.length())).append(")") .append(NEWLINE); output.append("Sequence "); - output.append(new Format("%" + maxid + "s").form(s2.getName())); - output.append(" : ").append(String.valueOf(s2.getStart())) - .append(" - ").append(String.valueOf(s2.getEnd())); + output.append(nameFormat.form(s2id)); output.append(" (Sequence length = ") .append(String.valueOf(s2str.length())).append(")") .append(NEWLINE).append(NEWLINE); @@ -508,7 +503,7 @@ public class AlignSeq for (int j = 0; j < nochunks; j++) { // Print the first aligned sequence - output.append(new Format("%" + (maxid) + "s").form(s1id)).append(" "); + output.append(nameFormat.form(s1id)).append(" "); for (int i = 0; i < len; i++) { @@ -519,7 +514,7 @@ public class AlignSeq } output.append(NEWLINE); - output.append(new Format("%" + (maxid) + "s").form(" ")).append(" "); + output.append(nameFormat.form(" ")).append(" "); /* * Print out the match symbols: @@ -539,7 +534,7 @@ public class AlignSeq pid++; output.append("|"); } - else if (type.equals("pep")) + else if (PEP.equals(type)) { if (pam250.getPairwiseScore(c1, c2) > 0) { @@ -559,8 +554,7 @@ public class AlignSeq // Now print the second aligned sequence output = output.append(NEWLINE); - output = output.append(new Format("%" + (maxid) + "s").form(s2id)) - .append(" "); + output = output.append(nameFormat.form(s2id)).append(" "); for (int i = 0; i < len; i++) { @@ -574,7 +568,8 @@ public class AlignSeq } pid = pid / (aseq1.length - count) * 100; - output = output.append(new Format("Percentage ID = %2.2f\n").form(pid)); + output.append(new Format("Percentage ID = %3.2f\n").form(pid)); + output.append(NEWLINE); try { os.print(output.toString()); @@ -596,7 +591,9 @@ public class AlignSeq public int findTrace(int i, int j) { int t = 0; - float max = score[i - 1][j - 1] + (lookup[seq1[i]][seq2[j]] * 10); + float pairwiseScore = scoreMatrix.getPairwiseScore(s1str.charAt(i), + s2str.charAt(j)); + float max = score[i - 1][j - 1] + (pairwiseScore * 10); if (F[i][j] > max) { @@ -640,18 +637,23 @@ public class AlignSeq int m = seq2.length; // top left hand element - score[0][0] = lookup[seq1[0]][seq2[0]] * 10; - E[0][0] = -gapExtend; + score[0][0] = scoreMatrix.getPairwiseScore(s1str.charAt(0), + s2str.charAt(0)) * 10; + E[0][0] = -GAP_EXTEND_COST; F[0][0] = 0; // Calculate the top row first for (int j = 1; j < m; j++) { // What should these values be? 0 maybe - E[0][j] = max(score[0][j - 1] - gapOpen, E[0][j - 1] - gapExtend); - F[0][j] = -gapExtend; + E[0][j] = max(score[0][j - 1] - GAP_OPEN_COST, + E[0][j - 1] - GAP_EXTEND_COST); + F[0][j] = -GAP_EXTEND_COST; - score[0][j] = max(lookup[seq1[0]][seq2[j]] * 10, -gapOpen, -gapExtend); + float pairwiseScore = scoreMatrix.getPairwiseScore(s1str.charAt(0), + s2str.charAt(j)); + score[0][j] = max(pairwiseScore * 10, -GAP_OPEN_COST, + -GAP_EXTEND_COST); traceback[0][j] = 1; } @@ -659,10 +661,13 @@ public class AlignSeq // Now do the left hand column for (int i = 1; i < n; i++) { - E[i][0] = -gapOpen; - F[i][0] = max(score[i - 1][0] - gapOpen, F[i - 1][0] - gapExtend); + E[i][0] = -GAP_OPEN_COST; + F[i][0] = max(score[i - 1][0] - GAP_OPEN_COST, + F[i - 1][0] - GAP_EXTEND_COST); - score[i][0] = max(lookup[seq1[i]][seq2[0]] * 10, E[i][0], F[i][0]); + float pairwiseScore = scoreMatrix.getPairwiseScore(s1str.charAt(i), + s2str.charAt(0)); + score[i][0] = max(pairwiseScore * 10, E[i][0], F[i][0]); traceback[i][0] = -1; } @@ -671,11 +676,15 @@ public class AlignSeq { for (int j = 1; j < m; j++) { - E[i][j] = max(score[i][j - 1] - gapOpen, E[i][j - 1] - gapExtend); - F[i][j] = max(score[i - 1][j] - gapOpen, F[i - 1][j] - gapExtend); - - score[i][j] = max(score[i - 1][j - 1] - + (lookup[seq1[i]][seq2[j]] * 10), E[i][j], F[i][j]); + E[i][j] = max(score[i][j - 1] - GAP_OPEN_COST, + E[i][j - 1] - GAP_EXTEND_COST); + F[i][j] = max(score[i - 1][j] - GAP_OPEN_COST, + F[i - 1][j] - GAP_EXTEND_COST); + + float pairwiseScore = scoreMatrix.getPairwiseScore(s1str.charAt(i), + s2str.charAt(j)); + score[i][j] = max(score[i - 1][j - 1] + (pairwiseScore * 10), + E[i][j], F[i][j]); traceback[i][j] = findTrace(i, j); } } @@ -774,7 +783,7 @@ public class AlignSeq for (int i = 0; i < s.length(); i++) { char c = s.charAt(i); - encoded[i] = scoreModel.getMatrixIndex(c); + encoded[i] = scoreMatrix.getMatrixIndex(c); } return encoded; @@ -797,7 +806,7 @@ public class AlignSeq public static void displayMatrix(Graphics g, int[][] mat, int n, int m, int psize) { - // TODO method dosen't seem to be referenced anywhere delete?? + // TODO method doesn't seem to be referenced anywhere delete?? int max = -1000; int min = 1000; @@ -861,7 +870,8 @@ public class AlignSeq */ public jalview.datamodel.Mapping getMappingFromS1(boolean allowmismatch) { - ArrayList as1 = new ArrayList(), as2 = new ArrayList(); + ArrayList as1 = new ArrayList(), + as2 = new ArrayList(); int pdbpos = s2.getStart() + getSeq2Start() - 2; int alignpos = s1.getStart() + getSeq1Start() - 2; int lp2 = pdbpos - 3, lp1 = alignpos - 3; @@ -905,8 +915,8 @@ public class AlignSeq } // construct range pairs - int[] mapseq1 = new int[as1.size() + (lastmatch ? 1 : 0)], mapseq2 = new int[as2 - .size() + (lastmatch ? 1 : 0)]; + int[] mapseq1 = new int[as1.size() + (lastmatch ? 1 : 0)], + mapseq2 = new int[as2.size() + (lastmatch ? 1 : 0)]; int i = 0; for (Integer ip : as1) { @@ -949,7 +959,8 @@ public class AlignSeq List ochains, AlignmentI al, String dnaOrProtein, boolean removeOldAnnots) { - List orig = new ArrayList(), repl = new ArrayList(); + List orig = new ArrayList(), + repl = new ArrayList(); List aligs = new ArrayList(); if (al != null && al.getHeight() > 0) { @@ -971,8 +982,8 @@ public class AlignSeq bestm = msq; } } - System.out.println("Best Score for " + (matches.size() + 1) + " :" - + bestscore); + // System.out.println("Best Score for " + (matches.size() + 1) + " :" + // + bestscore); matches.add(bestm); aligns.add(bestaseq); al.deleteSequence(bestm);