X-Git-Url: http://source.jalview.org/gitweb/?a=blobdiff_plain;ds=sidebyside;f=src%2Fjalview%2Fanalysis%2FAlignSeq.java;h=4ec0457c2980a7977245355c48bfc8d2cf1410cc;hb=6c36212c1e16557e6afb1b4dba9c28864d52ad4a;hp=a27c51bc71763c716df8a84769c682630227b6c7;hpb=f7f284d2b85186bb9133e7d9d0a245661ba349c8;p=jalview.git
diff --git a/src/jalview/analysis/AlignSeq.java b/src/jalview/analysis/AlignSeq.java
index a27c51b..4ec0457 100755
--- a/src/jalview/analysis/AlignSeq.java
+++ b/src/jalview/analysis/AlignSeq.java
@@ -1,29 +1,51 @@
/*
- * Jalview - A Sequence Alignment Editor and Viewer (Version 2.6)
- * Copyright (C) 2010 J Procter, AM Waterhouse, G Barton, M Clamp, S Searle
+ * Jalview - A Sequence Alignment Editor and Viewer ($$Version-Rel$$)
+ * Copyright (C) $$Year-Rel$$ The Jalview Authors
*
* This file is part of Jalview.
*
* Jalview is free software: you can redistribute it and/or
* modify it under the terms of the GNU General Public License
- * as published by the Free Software Foundation, either version 3 of the License, or (at your option) any later version.
- *
+ * as published by the Free Software Foundation, either version 3
+ * of the License, or (at your option) any later version.
+ *
* Jalview is distributed in the hope that it will be useful, but
* WITHOUT ANY WARRANTY; without even the implied warranty
* of MERCHANTABILITY or FITNESS FOR A PARTICULAR
* PURPOSE. See the GNU General Public License for more details.
*
- * You should have received a copy of the GNU General Public License along with Jalview. If not, see .
+ * You should have received a copy of the GNU General Public License
+ * along with Jalview. If not, see .
+ * The Jalview Authors are detailed in the 'AUTHORS' file.
*/
package jalview.analysis;
-import java.util.*;
-
-import java.awt.*;
-
-import jalview.datamodel.*;
-import jalview.schemes.*;
-import jalview.util.*;
+import java.util.Locale;
+
+import jalview.analysis.scoremodels.PIDModel;
+import jalview.analysis.scoremodels.ScoreMatrix;
+import jalview.analysis.scoremodels.ScoreModels;
+import jalview.analysis.scoremodels.SimilarityParams;
+import jalview.datamodel.AlignmentAnnotation;
+import jalview.datamodel.AlignmentI;
+import jalview.datamodel.Mapping;
+import jalview.datamodel.Sequence;
+import jalview.datamodel.SequenceI;
+import jalview.math.MiscMath;
+import jalview.util.Comparison;
+import jalview.util.Format;
+import jalview.util.MapList;
+import jalview.util.MessageManager;
+
+import java.awt.Color;
+import java.awt.Graphics;
+import java.io.PrintStream;
+import java.lang.IllegalArgumentException;
+import java.util.ArrayList;
+import java.util.Arrays;
+import java.util.HashMap;
+import java.util.List;
+import java.util.StringTokenizer;
/**
*
@@ -33,25 +55,32 @@ import jalview.util.*;
*/
public class AlignSeq
{
+ private static final int MAX_NAME_LENGTH = 30;
+
+ //&!
+ private static final int GAP_OPEN_COST = 120;
+ //private static final int GAP_OPEN_COST = 100;
+
+ private static final int GAP_EXTEND_COST = 20;
+ //private static final int GAP_EXTEND_COST = 5;
+
+ private static final int GAP_INDEX = -1;
+
public static final String PEP = "pep";
public static final String DNA = "dna";
- static String[] dna =
- { "A", "C", "G", "T", "-" };
+ private static final String NEWLINE = System.lineSeparator();
- // "C", "T", "A", "G", "-"};
- static String[] pep =
- { "A", "R", "N", "D", "C", "Q", "E", "G", "H", "I", "L", "K", "M", "F",
- "P", "S", "T", "W", "Y", "V", "B", "Z", "X", "-" };
+ float[][] score;
- int[][] score;
+ float alignmentScore;
- int[][] E;
+ float[][] E;
- int[][] F;
+ float[][] F;
- int[][] traceback;
+ int[][] traceback; // todo is this actually used?
int[] seq1;
@@ -77,6 +106,10 @@ public class AlignSeq
public String astr2 = "";
+ public String indelfreeAstr1 = "";
+
+ public String indelfreeAstr2 = "";
+
/** DOCUMENT ME!! */
public int seq1start;
@@ -86,47 +119,37 @@ public class AlignSeq
/** DOCUMENT ME!! */
public int seq2start;
- /** DOCUMENT ME!! */
public int seq2end;
int count;
- /** DOCUMENT ME!! */
- public int maxscore;
-
- float pid;
-
- int prev = 0;
-
- int gapOpen = 120;
-
- int gapExtend = 20;
+ public float maxscore;
- int[][] lookup = ResidueProperties.getBLOSUM62();
+ public float meanScore; //needed for PaSiMap
- String[] intToStr = pep;
+ public int hypotheticMaxScore; // needed for PaSiMap
- int defInt = 23;
+ int prev = 0;
StringBuffer output = new StringBuffer();
- String type;
+ String type; // AlignSeq.PEP or AlignSeq.DNA
- private int[] charToInt;
+ private ScoreMatrix scoreMatrix;
/**
* Creates a new AlignSeq object.
*
* @param s1
- * DOCUMENT ME!
+ * first sequence for alignment
* @param s2
- * DOCUMENT ME!
+ * second sequence for alignment
* @param type
- * DOCUMENT ME!
+ * molecule type, either AlignSeq.PEP or AlignSeq.DNA
*/
public AlignSeq(SequenceI s1, SequenceI s2, String type)
{
- SeqInit(s1, s1.getSequenceAsString(), s2, s2.getSequenceAsString(),
+ seqInit(s1, s1.getSequenceAsString(), s2, s2.getSequenceAsString(),
type);
}
@@ -143,7 +166,8 @@ 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);
}
/**
@@ -151,12 +175,22 @@ public class AlignSeq
*
* @return DOCUMENT ME!
*/
- public int getMaxScore()
+ public float getMaxScore()
{
return maxscore;
}
/**
+ * returns the overall score of the alignment
+ *
+ * @return
+ */
+ public float getAlignmentScore()
+ {
+ return alignmentScore;
+ }
+
+ /**
* DOCUMENT ME!
*
* @return DOCUMENT ME!
@@ -247,55 +281,54 @@ public class AlignSeq
}
/**
- * DOCUMENT ME!
*
- * @return DOCUMENT ME!
+ * @return aligned instance of Seq 1
*/
- public SequenceI getS1()
+ public SequenceI getAlignedSeq1()
{
- return s1;
+ 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());
+ return alSeq1;
}
/**
- * DOCUMENT ME!
*
- * @return DOCUMENT ME!
+ * @return aligned instance of Seq 2
*/
- public SequenceI getS2()
+ public SequenceI getAlignedSeq2()
{
- return s2;
+ 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());
+ return alSeq2;
}
/**
- * DOCUMENT ME!
+ * Construct score matrix for sequences with standard DNA or PEPTIDE matrix
*
* @param s1
- * DOCUMENT ME!
+ * - sequence 1
* @param string1
- * - string to align for sequence1
+ * - string to use for s1
* @param s2
- * sequence 2
+ * - sequence 2
* @param string2
- * - string to align for sequence2
+ * - string to use for s2
* @param type
* DNA or PEPTIDE
*/
- public void SeqInit(SequenceI s1, String string1, SequenceI s2,
+ public void seqInit(SequenceI s1, String string1, SequenceI s2,
String string2, String type)
{
this.s1 = s1;
this.s2 = s2;
setDefaultParams(type);
- SeqInit(string1, string2);
- }
-
- public void SeqInit(SequenceI s1, String string1, SequenceI s2,
- String string2, ScoreMatrix scoreMatrix)
- {
- this.s1 = s1;
- this.s2 = s2;
- setType(scoreMatrix.isDNA() ? AlignSeq.DNA : AlignSeq.PEP);
- lookup = scoreMatrix.getMatrix();
+ seqInit(string1, string2);
}
/**
@@ -305,106 +338,150 @@ public class AlignSeq
* @param string1
* @param string2
*/
- private void SeqInit(String string1, String string2)
+ private void seqInit(String string1, String string2)
{
s1str = extractGaps(jalview.util.Comparison.GapChars, string1);
s2str = extractGaps(jalview.util.Comparison.GapChars, string2);
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;
}
- // System.out.println("lookuip " + rt.freeMemory() + " "+ rt.totalMemory());
- seq1 = new int[s1str.length()];
-
- // System.out.println("seq1 " + rt.freeMemory() +" " + rt.totalMemory());
- seq2 = new int[s2str.length()];
+ score = new float[s1str.length()][s2str.length()];
- // System.out.println("seq2 " + rt.freeMemory() + " " + rt.totalMemory());
- score = new int[s1str.length()][s2str.length()];
+ E = new float[s1str.length()][s2str.length()];
- // System.out.println("score " + rt.freeMemory() + " " + rt.totalMemory());
- E = new int[s1str.length()][s2str.length()];
-
- // System.out.println("E " + rt.freeMemory() + " " + rt.totalMemory());
- F = new int[s1str.length()][s2str.length()];
+ F = new float[s1str.length()][s2str.length()];
traceback = new int[s1str.length()][s2str.length()];
- // System.out.println("F " + rt.freeMemory() + " " + rt.totalMemory());
- seq1 = stringToInt(s1str, type);
+ seq1 = indexEncode(s1str);
+
+ seq2 = indexEncode(s2str);
+ }
- // System.out.println("seq1 " + rt.freeMemory() + " " + rt.totalMemory());
- seq2 = stringToInt(s2str, type);
+ private void setDefaultParams(String moleculeType)
+ {
+ 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 }));
+ }
- // System.out.println("Seq2 " + rt.freeMemory() + " " + rt.totalMemory());
- // long tstart = System.currentTimeMillis();
- // calcScoreMatrix();
- // long tend = System.currentTimeMillis();
- // System.out.println("Time take to calculate score matrix = " +
- // (tend-tstart) + " ms");
- // printScoreMatrix(score);
- // System.out.println();
- // printScoreMatrix(traceback);
- // System.out.println();
- // printScoreMatrix(E);
- // System.out.println();
- // /printScoreMatrix(F);
- // System.out.println();
- // tstart = System.currentTimeMillis();
- // traceAlignment();
- // tend = System.currentTimeMillis();
- // System.out.println("Time take to traceback alignment = " + (tend-tstart)
- // + " ms");
+ type = moleculeType;
+ scoreMatrix = ScoreModels.getInstance()
+ .getDefaultModel(PEP.equals(type));
}
- private void setDefaultParams(String type)
+ /**
+ * DOCUMENT ME!
+ */
+ public void traceAlignment()
{
- setType(type);
+ // Find the maximum score along the rhs or bottom row
+ float max = -Float.MAX_VALUE;
- if (type.equals(AlignSeq.PEP))
+ for (int i = 0; i < seq1.length; i++)
{
- lookup = ResidueProperties.getDefaultPeptideMatrix();
+ if (score[i][seq2.length - 1] > max)
+ {
+ max = score[i][seq2.length - 1];
+ maxi = i;
+ maxj = seq2.length - 1;
+ }
}
- else if (type.equals(AlignSeq.DNA))
+
+ for (int j = 0; j < seq2.length; j++)
{
- lookup = ResidueProperties.getDefaultDnaMatrix();
+ if (score[seq1.length - 1][j] > max)
+ {
+ max = score[seq1.length - 1][j];
+ maxi = seq1.length - 1;
+ maxj = j;
+ }
}
- }
- private void setType(String type2)
- {
- this.type = type2;
- if (type.equals(AlignSeq.PEP))
+ int i = maxi;
+ int j = maxj;
+ int trace;
+ maxscore = score[i][j] / 10f;
+
+
+ aseq1 = new int[seq1.length + seq2.length];
+ aseq2 = new int[seq1.length + seq2.length];
+
+ StringBuilder sb1 = new StringBuilder(aseq1.length);
+ StringBuilder sb2 = new StringBuilder(aseq2.length);
+
+ count = (seq1.length + seq2.length) - 1;
+
+
+ while (i > 0 && j > 0)
{
- intToStr = pep;
- charToInt = ResidueProperties.aaIndex;
- defInt = 23;
+ aseq1[count] = seq1[i];
+ sb1.append(s1str.charAt(i));
+ aseq2[count] = seq2[j];
+ sb2.append(s2str.charAt(j));
+
+ trace = findTrace(i, j);
+
+ if (trace == 0)
+ {
+ i--;
+ j--;
+ }
+ else if (trace == 1)
+ {
+ j--;
+ aseq1[count] = GAP_INDEX;
+ sb1.replace(sb1.length() - 1, sb1.length(), "-");
+ }
+ else if (trace == -1)
+ {
+ i--;
+ aseq2[count] = GAP_INDEX;
+ sb2.replace(sb2.length() - 1, sb2.length(), "-");
+ }
+
+ count--;
}
- else if (type.equals(AlignSeq.DNA))
+
+ seq1start = i + 1;
+ seq2start = j + 1;
+
+ if (aseq1[count] != GAP_INDEX)
{
- intToStr = dna;
- charToInt = ResidueProperties.nucleotideIndex;
- defInt = 4;
+ aseq1[count] = seq1[i];
+ sb1.append(s1str.charAt(i));
}
- else
+
+ if (aseq2[count] != GAP_INDEX)
{
- output.append("Wrong type = dna or pep only");
- throw new Error("Unknown Type " + type2
- + " - dna or pep are the only allowed values.");
+ aseq2[count] = seq2[j];
+ sb2.append(s2str.charAt(j));
}
+
+
+ /*
+ * we built the character strings backwards, so now
+ * reverse them to convert to sequence strings
+ */
+ astr1 = sb1.reverse().toString();
+ astr2 = sb2.reverse().toString();
}
/**
* DOCUMENT ME!
*/
- public void traceAlignment()
+ public void traceAlignmentWithEndGaps()
{
// Find the maximum score along the rhs or bottom row
- int max = -9999;
+ float max = -Float.MAX_VALUE;
for (int i = 0; i < seq1.length; i++)
{
@@ -426,33 +503,55 @@ 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;
+
+ //&! get trailing gaps
+ while ((i < seq1.length - 1) || (j < seq2.length - 1))
+ {
+ i++;
+ j++;
+ }
+ seq1end = i + 1;
+ seq2end = j + 1;
- seq1end = maxi + 1;
- seq2end = maxj + 1;
aseq1 = new int[seq1.length + seq2.length];
aseq2 = new int[seq1.length + seq2.length];
+ StringBuilder sb1 = new StringBuilder(aseq1.length);
+ StringBuilder sb2 = new StringBuilder(aseq2.length);
+
count = (seq1.length + seq2.length) - 1;
- while ((i > 0) && (j > 0))
+ //&! get trailing gaps
+ while ((i >= seq1.length) || (j >= seq2.length))
{
- if ((aseq1[count] != defInt) && (i >= 0))
+ if (i >= seq1.length)
{
- aseq1[count] = seq1[i];
- astr1 = s1str.charAt(i) + astr1;
+ aseq1[count] = GAP_INDEX;
+ sb1.append("-");
+ aseq2[count] = seq2[j];
+ sb2.append(s2str.charAt(j));
+ } else if (j >= seq2.length) {
+ aseq1[count] = seq1[i];
+ sb1.append(s1str.charAt(i));
+ aseq2[count] = GAP_INDEX;
+ sb2.append("-");
}
+ i--;
+ j--;
+ }
- if ((aseq2[count] != defInt) && (j > 0))
- {
- aseq2[count] = seq2[j];
- astr2 = s2str.charAt(j) + astr2;
- }
+
+ while (i > 0 && j > 0)
+ {
+ aseq1[count] = seq1[i];
+ sb1.append(s1str.charAt(i));
+ aseq2[count] = seq2[j];
+ sb2.append(s2str.charAt(j));
trace = findTrace(i, j);
@@ -464,14 +563,14 @@ public class AlignSeq
else if (trace == 1)
{
j--;
- aseq1[count] = defInt;
- astr1 = "-" + astr1.substring(1);
+ aseq1[count] = GAP_INDEX;
+ sb1.replace(sb1.length() - 1, sb1.length(), "-");
}
else if (trace == -1)
{
i--;
- aseq2[count] = defInt;
- astr2 = "-" + astr2.substring(1);
+ aseq2[count] = GAP_INDEX;
+ sb2.replace(sb2.length() - 1, sb2.length(), "-");
}
count--;
@@ -480,65 +579,96 @@ public class AlignSeq
seq1start = i + 1;
seq2start = j + 1;
- if (aseq1[count] != defInt)
+ if (aseq1[count] != GAP_INDEX)
{
aseq1[count] = seq1[i];
- astr1 = s1str.charAt(i) + astr1;
+ sb1.append(s1str.charAt(i));
}
- if (aseq2[count] != defInt)
+ if (aseq2[count] != GAP_INDEX)
{
aseq2[count] = seq2[j];
- astr2 = s2str.charAt(j) + astr2;
+ sb2.append(s2str.charAt(j));
+ }
+
+ //&! get initial gaps
+ while (j > 0 || i > 0)
+ {
+ if (j > 0)
+ {
+ sb1.append("-");
+ sb2.append(s2str.charAt(j));
+ j--;
+ } else if (i > 0) {
+ sb1.append(s1str.charAt(i));
+ sb2.append("-");
+ i--;
+ }
}
+
+ /*
+ * we built the character strings backwards, so now
+ * reverse them to convert to sequence strings
+ */
+ astr1 = sb1.reverse().toString();
+ astr2 = sb2.reverse().toString();
}
/**
* 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)
- {
- maxid = s2.getName().length();
- }
- if (maxid > 30)
+ String s1id = getAlignedSeq1().getDisplayId(true);
+ String s2id = getAlignedSeq2().getDisplayId(true);
+ int nameLength = Math.max(s1id.length(), s2id.length());
+ if (nameLength > MAX_NAME_LENGTH)
{
- 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 nochunks = ((aseq1.length - count) / len) + 1;
- pid = 0;
-
- output.append("Score = " + score[maxi][maxj] + "\n");
- output.append("Length of alignment = " + (aseq1.length - count) + "\n");
+ int len = 72 - nameLength - 1;
+ int nochunks = ((aseq1.length - count) / len)
+ + ((aseq1.length - count) % len > 0 ? 1 : 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(" : " + s1.getStart() + " - " + s1.getEnd()
- + " (Sequence length = " + s1str.length() + ")\n");
+ 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(" : " + s2.getStart() + " - " + s2.getEnd()
- + " (Sequence length = " + s2str.length() + ")\n\n");
+ output.append(nameFormat.form(s2id));
+ output.append(" (Sequence length = ")
+ .append(String.valueOf(s2str.length())).append(")")
+ .append(NEWLINE).append(NEWLINE);
+
+ ScoreMatrix pam250 = ScoreModels.getInstance().getPam250();
for (int j = 0; j < nochunks; j++)
{
// Print the first aligned sequence
- output.append(new Format("%" + (maxid) + "s").form(s1id) + " ");
+ output.append(nameFormat.form(s1id)).append(" ");
for (int i = 0; i < len; i++)
{
@@ -548,25 +678,30 @@ public class AlignSeq
}
}
- output.append("\n");
- output.append(new Format("%" + (maxid) + "s").form(" ") + " ");
+ output.append(NEWLINE);
+ output.append(nameFormat.form(" ")).append(" ");
- // Print out the matching chars
+ /*
+ * Print out the match symbols:
+ * | for exact match (ignoring case)
+ * . if PAM250 score is positive
+ * else a space
+ */
for (int i = 0; i < len; i++)
{
if ((i + (j * len)) < astr1.length())
{
- if (astr1.charAt(i + (j * len)) == astr2.charAt(i + (j * len))
- && !jalview.util.Comparison.isGap(astr1.charAt(i
- + (j * len))))
+ char c1 = astr1.charAt(i + (j * len));
+ char c2 = astr2.charAt(i + (j * len));
+ boolean sameChar = Comparison.isSameResidue(c1, c2, false);
+ if (sameChar && !Comparison.isGap(c1))
{
pid++;
output.append("|");
}
- else if (type.equals("pep"))
+ else if (PEP.equals(type))
{
- if (ResidueProperties.getPAM250(astr1.charAt(i + (j * len)),
- astr2.charAt(i + (j * len))) > 0)
+ if (pam250.getPairwiseScore(c1, c2) > 0)
{
output.append(".");
}
@@ -583,9 +718,8 @@ public class AlignSeq
}
// Now print the second aligned sequence
- output = output.append("\n");
- output = output.append(new Format("%" + (maxid) + "s").form(s2id)
- + " ");
+ output = output.append(NEWLINE);
+ output = output.append(nameFormat.form(s2id)).append(" ");
for (int i = 0; i < len; i++)
{
@@ -595,13 +729,12 @@ public class AlignSeq
}
}
- output = output.append("\n\n");
+ output.append(NEWLINE).append(NEWLINE);
}
- pid = pid / (float) (aseq1.length - count) * 100;
- output = output.append(new Format("Percentage ID = %2.2f\n\n")
- .form(pid));
-
+ pid = pid / (aseq1.length - count) * 100;
+ output.append(new Format("Percentage ID = %3.2f\n").form(pid));
+ output.append(NEWLINE);
try
{
os.print(output.toString());
@@ -613,46 +746,6 @@ public class AlignSeq
/**
* DOCUMENT ME!
*
- * @param mat
- * DOCUMENT ME!
- */
- public void printScoreMatrix(int[][] mat)
- {
- int n = seq1.length;
- int m = seq2.length;
-
- for (int i = 0; i < n; i++)
- {
- // Print the top sequence
- if (i == 0)
- {
- Format.print(System.out, "%8s", s2str.substring(0, 1));
-
- for (int jj = 1; jj < m; jj++)
- {
- Format.print(System.out, "%5s", s2str.substring(jj, jj + 1));
- }
-
- System.out.println();
- }
-
- for (int j = 0; j < m; j++)
- {
- if (j == 0)
- {
- Format.print(System.out, "%3s", s1str.substring(i, i + 1));
- }
-
- Format.print(System.out, "%3d ", mat[i][j] / 10);
- }
-
- System.out.println();
- }
- }
-
- /**
- * DOCUMENT ME!
- *
* @param i
* DOCUMENT ME!
* @param j
@@ -663,7 +756,9 @@ public class AlignSeq
public int findTrace(int i, int j)
{
int t = 0;
- int 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)
{
@@ -707,18 +802,23 @@ 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;
}
@@ -726,10 +826,13 @@ 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;
}
@@ -738,30 +841,38 @@ 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);
}
}
}
/**
- * DOCUMENT ME!
+ * Returns the given sequence with all of the given gap characters removed.
*
- * @param gapChar
- * DOCUMENT ME!
+ * @param gapChars
+ * a string of characters to be treated as gaps
* @param seq
- * DOCUMENT ME!
+ * the input sequence
*
- * @return DOCUMENT ME!
+ * @return
*/
- public static String extractGaps(String gapChar, String seq)
+ public static String extractGaps(String gapChars, String seq)
{
- StringTokenizer str = new StringTokenizer(seq, gapChar);
- StringBuffer newString = new StringBuffer();
+ if (gapChars == null || seq == null)
+ {
+ return null;
+ }
+ StringTokenizer str = new StringTokenizer(seq, gapChars);
+ StringBuilder newString = new StringBuilder(seq.length());
while (str.hasMoreTokens())
{
@@ -774,27 +885,27 @@ public class AlignSeq
/**
* DOCUMENT ME!
*
- * @param i1
+ * @param f1
* DOCUMENT ME!
- * @param i2
+ * @param f2
* DOCUMENT ME!
- * @param i3
+ * @param f3
* DOCUMENT ME!
*
* @return DOCUMENT ME!
*/
- public int max(int i1, int i2, int i3)
+ private static float max(float f1, float f2, float f3)
{
- int max = i1;
+ float max = f1;
- if (i2 > i1)
+ if (f2 > f1)
{
- max = i2;
+ max = f2;
}
- if (i3 > max)
+ if (f3 > max)
{
- max = i3;
+ max = f3;
}
return max;
@@ -803,65 +914,44 @@ public class AlignSeq
/**
* DOCUMENT ME!
*
- * @param i1
+ * @param f1
* DOCUMENT ME!
- * @param i2
+ * @param f2
* DOCUMENT ME!
*
* @return DOCUMENT ME!
*/
- public int max(int i1, int i2)
+ private static float max(float f1, float f2)
{
- int max = i1;
+ float max = f1;
- if (i2 > i1)
+ if (f2 > f1)
{
- max = i2;
+ max = f2;
}
return max;
}
/**
- * DOCUMENT ME!
+ * Converts the character string to an array of integers which are the
+ * corresponding indices to the characters in the score matrix
*
* @param s
- * DOCUMENT ME!
- * @param type
- * DOCUMENT ME!
*
- * @return DOCUMENT ME!
+ * @return
*/
- public int[] stringToInt(String s, String type)
+ int[] indexEncode(String s)
{
- int[] seq1 = new int[s.length()];
+ int[] encoded = new int[s.length()];
for (int i = 0; i < s.length(); i++)
{
- // String ss = s.substring(i, i + 1).toUpperCase();
char c = s.charAt(i);
- if ('a' <= c && c <= 'z')
- {
- // TO UPPERCASE !!!
- c -= ('a' - 'A');
- }
-
- try
- {
- seq1[i] = charToInt[c]; // set accordingly from setType
- if (seq1[i] < 0 || seq1[i] > defInt) // set from setType: 23 for
- // peptides, or 4 for NA.
- {
- seq1[i] = defInt;
- }
-
- } catch (Exception e)
- {
- seq1[i] = defInt;
- }
+ encoded[i] = scoreMatrix.getMatrixIndex(c);
}
- return seq1;
+ return encoded;
}
/**
@@ -881,6 +971,7 @@ public class AlignSeq
public static void displayMatrix(Graphics g, int[][] mat, int n, int m,
int psize)
{
+ // TODO method doesn't seem to be referenced anywhere delete??
int max = -1000;
int min = 1000;
@@ -918,4 +1009,434 @@ public class AlignSeq
}
}
}
+
+ /**
+ * Compute a globally optimal needleman and wunsch alignment between two
+ * sequences
+ *
+ * @param s1
+ * @param s2
+ * @param type
+ * AlignSeq.DNA or AlignSeq.PEP
+ */
+ public static AlignSeq doGlobalNWAlignment(SequenceI s1, SequenceI s2,
+ String type)
+ {
+ AlignSeq as = new AlignSeq(s1, s2, type);
+
+ as.calcScoreMatrix();
+ as.traceAlignment();
+ return as;
+ }
+
+ /**
+ *
+ * @return mapping from positions in S1 to corresponding positions in S2
+ */
+ public jalview.datamodel.Mapping getMappingFromS1(boolean allowmismatch)
+ {
+ 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;
+ boolean lastmatch = false;
+ // and now trace the alignment onto the atom set.
+ for (int i = 0; i < astr1.length(); i++)
+ {
+ char c1 = astr1.charAt(i), c2 = astr2.charAt(i);
+ if (c1 != '-')
+ {
+ alignpos++;
+ }
+
+ if (c2 != '-')
+ {
+ pdbpos++;
+ }
+
+ // ignore case differences
+ if (allowmismatch || (c1 == c2) || (Math.abs(c2-c1)==('a'-'A')))
+ {
+ // extend mapping interval
+ if (lp1 + 1 != alignpos || lp2 + 1 != pdbpos)
+ {
+ as1.add(Integer.valueOf(alignpos));
+ as2.add(Integer.valueOf(pdbpos));
+ }
+ lastmatch = true;
+ lp1 = alignpos;
+ lp2 = pdbpos;
+ }
+ else
+ {
+ // extend mapping interval
+ if (lastmatch)
+ {
+ as1.add(Integer.valueOf(lp1));
+ as2.add(Integer.valueOf(lp2));
+ }
+ lastmatch = false;
+ }
+ }
+ // construct range pairs
+
+ int[] mapseq1 = new int[as1.size() + (lastmatch ? 1 : 0)],
+ mapseq2 = new int[as2.size() + (lastmatch ? 1 : 0)];
+ int i = 0;
+ for (Integer ip : as1)
+ {
+ mapseq1[i++] = ip;
+ }
+ ;
+ i = 0;
+ for (Integer ip : as2)
+ {
+ mapseq2[i++] = ip;
+ }
+ ;
+ if (lastmatch)
+ {
+ mapseq1[mapseq1.length - 1] = alignpos;
+ mapseq2[mapseq2.length - 1] = pdbpos;
+ }
+ MapList map = new MapList(mapseq1, mapseq2, 1, 1);
+
+ jalview.datamodel.Mapping mapping = new Mapping(map);
+ mapping.setTo(s2);
+ return mapping;
+ }
+
+ /**
+ * matches ochains against al and populates seqs with the best match between
+ * each ochain and the set in al
+ *
+ * @param ochains
+ * @param al
+ * @param dnaOrProtein
+ * @param removeOldAnnots
+ * when true, old annotation is cleared before new annotation
+ * transferred
+ * @return List originals, List replacement,
+ * List alignment between each>
+ */
+ public static List> replaceMatchingSeqsWith(
+ List seqs, List annotations,
+ List ochains, AlignmentI al, String dnaOrProtein,
+ boolean removeOldAnnots)
+ {
+ List orig = new ArrayList(),
+ repl = new ArrayList();
+ List aligs = new ArrayList();
+ if (al != null && al.getHeight() > 0)
+ {
+ ArrayList matches = new ArrayList();
+ ArrayList aligns = new ArrayList();
+
+ for (SequenceI sq : ochains)
+ {
+ SequenceI bestm = null;
+ AlignSeq bestaseq = null;
+ float bestscore = 0;
+ for (SequenceI msq : al.getSequences())
+ {
+ AlignSeq aseq = doGlobalNWAlignment(msq, sq, dnaOrProtein);
+ if (bestm == null || aseq.getMaxScore() > bestscore)
+ {
+ bestscore = aseq.getMaxScore();
+ bestaseq = aseq;
+ bestm = msq;
+ }
+ }
+ // System.out.println("Best Score for " + (matches.size() + 1) + " :"
+ // + bestscore);
+ matches.add(bestm);
+ aligns.add(bestaseq);
+ al.deleteSequence(bestm);
+ }
+ for (int p = 0, pSize = seqs.size(); p < pSize; p++)
+ {
+ SequenceI sq, sp = seqs.get(p);
+ int q;
+ if ((q = ochains.indexOf(sp)) > -1)
+ {
+ seqs.set(p, sq = matches.get(q));
+ orig.add(sp);
+ repl.add(sq);
+ sq.setName(sp.getName());
+ sq.setDescription(sp.getDescription());
+ Mapping sp2sq;
+ sq.transferAnnotation(sp,
+ sp2sq = aligns.get(q).getMappingFromS1(false));
+ aligs.add(aligns.get(q));
+ int inspos = -1;
+ for (int ap = 0; ap < annotations.size();)
+ {
+ if (annotations.get(ap).sequenceRef == sp)
+ {
+ if (inspos == -1)
+ {
+ inspos = ap;
+ }
+ if (removeOldAnnots)
+ {
+ annotations.remove(ap);
+ }
+ else
+ {
+ AlignmentAnnotation alan = annotations.remove(ap);
+ alan.liftOver(sq, sp2sq);
+ alan.setSequenceRef(sq);
+ sq.addAlignmentAnnotation(alan);
+ }
+ }
+ else
+ {
+ ap++;
+ }
+ }
+ if (sq.getAnnotation() != null && sq.getAnnotation().length > 0)
+ {
+ annotations.addAll(inspos == -1 ? annotations.size() : inspos,
+ Arrays.asList(sq.getAnnotation()));
+ }
+ }
+ }
+ }
+ return Arrays.asList(orig, repl, aligs);
+ }
+
+ /**
+ * compute the PID vector used by the redundancy filter.
+ *
+ * @param originalSequences
+ * - sequences in alignment that are to filtered
+ * @param omitHidden
+ * - null or strings to be analysed (typically, visible portion of
+ * each sequence in alignment)
+ * @param start
+ * - first column in window for calculation
+ * @param end
+ * - last column in window for calculation
+ * @param ungapped
+ * - if true then use ungapped sequence to compute PID
+ * @return vector containing maximum PID for i-th sequence and any sequences
+ * longer than that seuqence
+ */
+ public static float[] computeRedundancyMatrix(
+ SequenceI[] originalSequences, String[] omitHidden, int start,
+ int end, boolean ungapped)
+ {
+ int height = originalSequences.length;
+ float[] redundancy = new float[height];
+ int[] lngth = new int[height];
+ for (int i = 0; i < height; i++)
+ {
+ redundancy[i] = 0f;
+ lngth[i] = -1;
+ }
+
+ // long start = System.currentTimeMillis();
+
+ SimilarityParams pidParams = new SimilarityParams(true, true, true,
+ true);
+ float pid;
+ String seqi, seqj;
+ for (int i = 0; i < height; i++)
+ {
+
+ for (int j = 0; j < i; j++)
+ {
+ if (i == j)
+ {
+ continue;
+ }
+
+ if (omitHidden == null)
+ {
+ seqi = originalSequences[i].getSequenceAsString(start, end);
+ seqj = originalSequences[j].getSequenceAsString(start, end);
+ }
+ else
+ {
+ seqi = omitHidden[i];
+ seqj = omitHidden[j];
+ }
+ if (lngth[i] == -1)
+ {
+ String ug = AlignSeq.extractGaps(Comparison.GapChars, seqi);
+ lngth[i] = ug.length();
+ if (ungapped)
+ {
+ seqi = ug;
+ }
+ }
+ if (lngth[j] == -1)
+ {
+ String ug = AlignSeq.extractGaps(Comparison.GapChars, seqj);
+ lngth[j] = ug.length();
+ if (ungapped)
+ {
+ seqj = ug;
+ }
+ }
+ pid = (float) PIDModel.computePID(seqi, seqj, pidParams);
+
+ // use real sequence length rather than string length
+ if (lngth[j] < lngth[i])
+ {
+ redundancy[j] = Math.max(pid, redundancy[j]);
+ }
+ else
+ {
+ redundancy[i] = Math.max(pid, redundancy[i]);
+ }
+
+ }
+ }
+ return redundancy;
+ }
+
+ /**
+ * calculate the mean score of the alignment
+ * mean score is equal to the score of an alignmenet of two sequences with randomly shuffled AA sequence composited of the same AA as the two original sequences
+ *
+ */
+ public void meanScore()
+ {
+ //int length = (indelfreeAstr1.length() > indelfreeAstr2.length()) ? indelfreeAstr1.length() : indelfreeAstr2.length();
+ int length = indelfreeAstr1.length(); //both have the same length
+ //create HashMap for counting residues in each sequence
+ HashMap seq1ResCount = new HashMap();
+ HashMap seq2ResCount = new HashMap();
+
+ // for both sequences (String indelfreeAstr1 or 2) create a key for the residue and add 1 each time its encountered
+ for (char residue: indelfreeAstr1.toCharArray())
+ {
+ seq1ResCount.putIfAbsent(residue, 0);
+ seq1ResCount.replace(residue, seq1ResCount.get(residue) + 1);
+ }
+ for (char residue: indelfreeAstr2.toCharArray())
+ {
+ seq2ResCount.putIfAbsent(residue, 0);
+ seq2ResCount.replace(residue, seq2ResCount.get(residue) + 1);
+ }
+
+ // meanscore = for each residue pair get the number of appearance and add (countA * countB * pairwiseScore(AB))
+ // divide the meanscore by the sequence length afterwards
+ float _meanscore = 0;
+ for (char resA : seq1ResCount.keySet())
+ {
+ for (char resB : seq2ResCount.keySet())
+ {
+ int countA = seq1ResCount.get(resA);
+ int countB = seq2ResCount.get(resB);
+
+ float scoreAB = scoreMatrix.getPairwiseScore(resA, resB);
+
+ _meanscore += countA * countB * scoreAB;
+ }
+ }
+ _meanscore /= length;
+ this.meanScore = _meanscore;
+ }
+
+ public float getMeanScore()
+ {
+ return this.meanScore;
+ }
+
+ /**
+ * calculate the hypothetic max score using the self-alignment of the sequences
+ */
+ public void hypotheticMaxScore()
+ {
+ int _hmsA = 0;
+ int _hmsB = 0;
+ for (char residue: indelfreeAstr1.toCharArray())
+ {
+ _hmsA += scoreMatrix.getPairwiseScore(residue, residue);
+ }
+ for (char residue: indelfreeAstr2.toCharArray())
+ {
+ _hmsB += scoreMatrix.getPairwiseScore(residue, residue);
+ }
+ this.hypotheticMaxScore = (_hmsA < _hmsB) ? _hmsA : _hmsB; // take the lower self alignment
+
+ }
+
+ public int getHypotheticMaxScore()
+ {
+ return this.hypotheticMaxScore;
+ }
+
+ /**
+ * create strings based of astr1 and astr2 but without gaps
+ */
+ public void getIndelfreeAstr()
+ {
+ int n = astr1.length(); // both have the same length
+ for (int i = 0; i < n; i++)
+ {
+ if (Character.isLetter(astr1.charAt(i)) && Character.isLetter(astr2.charAt(i))) // if both sequences dont have a gap -> add to indelfreeAstr
+ {
+ this.indelfreeAstr1 += astr1.charAt(i);
+ this.indelfreeAstr2 += astr2.charAt(i);
+ }
+ }
+ }
+
+ /**
+ * calculates the overall score of the alignment
+ * preprescore = sum of all scores - all penalties
+ * if preprescore < 1 ~ alignmentScore = Float.NaN >
+ * alignmentScore = ((preprescore - meanScore) / (hypotheticMaxScore - meanScore)) * coverage
+ */
+ public void scoreAlignment() throws RuntimeException
+ {
+
+ getIndelfreeAstr();
+ meanScore();
+ hypotheticMaxScore();
+ // cannot calculate score because denominator would be zero
+ if (this.hypotheticMaxScore == this.meanScore)
+ {
+ throw new IllegalArgumentException(String.format("hypotheticMaxScore (%8.2f) == meanScore (%8.2f) - division by 0", hypotheticMaxScore, meanScore));
+ }
+ //int n = (astr1.length() > astr2.length()) ? astr1.length() : astr2.length();
+ int n = indelfreeAstr1.length();
+
+ float score = 0;
+ boolean aGapOpen = false;
+ boolean bGapOpen = false;
+ for (int i = 0; i < n; i++)
+ {
+ char char1 = indelfreeAstr1.charAt(i);
+ char char2 = indelfreeAstr2.charAt(i);
+ boolean aIsLetter = Character.isLetter(char1);
+ boolean bIsLetter = Character.isLetter(char2);
+ if (aIsLetter && bIsLetter) // if pair -> get score
+ {
+ score += scoreMatrix.getPairwiseScore(char1, char2);
+ } else if (!aIsLetter && !bIsLetter) { // both are gap -> skip
+ } else if ((!aIsLetter && aGapOpen) || (!bIsLetter && bGapOpen)) { // one side gapopen -> score - gap_extend
+ score -= GAP_EXTEND_COST;
+ } else { // no gap open -> score - gap_open
+ score -= GAP_OPEN_COST;
+ }
+ // adjust GapOpen status in both sequences
+ aGapOpen = (!aIsLetter) ? true : false;
+ bGapOpen = (!bIsLetter) ? true : false;
+ }
+
+ float preprescore = score; // if this score < 1 --> alignment score = Float.NaN
+ score = (score - this.meanScore) / (this.hypotheticMaxScore - this.meanScore);
+ int[] _max = MiscMath.findMax(new int[]{astr1.replace("-","").length(), astr2.replace("-","").length()}); // {index of max, max}
+ float coverage = (float) n / (float) _max[1]; // indelfreeAstr length / longest sequence length
+ float prescore = score; // only debug
+ score *= coverage;
+
+ System.out.println(String.format("prepre-score: %f, pre-score: %f, longlength: %d\nscore: %f, mean: %f, max: %d", preprescore, prescore, _max[1], score, this.meanScore, this.hypotheticMaxScore));
+ this.alignmentScore = (preprescore < 1) ? Float.NaN : score;
+ }
}