"//";
private static final String AA_SEQS_1 =
- ">Seq1Name\n" +
+ ">Seq1Name/5-8\n" +
"K-QY--L\n" +
- ">Seq2Name\n" +
+ ">Seq2Name/12-15\n" +
"-R-FP-W-\n";
private static final String CDNA_SEQS_1 =
- ">Seq1Name\n" +
+ ">Seq1Name/100-111\n" +
"AC-GG--CUC-CAA-CT\n" +
- ">Seq2Name\n" +
+ ">Seq2Name/200-211\n" +
"-CG-TTA--ACG---AAGT\n";
private static final String CDNA_SEQS_2 =
- ">Seq1Name\n" +
+ ">Seq1Name/50-61\n" +
"GCTCGUCGTACT\n" +
- ">Seq2Name\n" +
+ ">Seq2Name/60-71\n" +
"GGGTCAGGCAGT\n";
// @formatter:on
* Make mappings between sequences. The 'aligned cDNA' is playing the role
* of what would normally be protein here.
*/
- AlignedCodonFrame acf = new AlignedCodonFrame();
- MapList ml = new MapList(new int[] { 1, 12 }, new int[] { 1, 12 }, 1, 1);
- acf.addMap(al2.getSequenceAt(0), al1.getSequenceAt(0), ml);
- acf.addMap(al2.getSequenceAt(1), al1.getSequenceAt(1), ml);
- al1.addCodonFrame(acf);
+ makeMappings(al2, al1);
((Alignment) al2).alignAs(al1, false, true);
assertEquals("GC-TC--GUC-GTA-CT", al2.getSequenceAt(0)
// see also AlignmentUtilsTests
AlignmentI al1 = loadAlignment(CDNA_SEQS_1, "FASTA");
AlignmentI al2 = loadAlignment(AA_SEQS_1, "FASTA");
- AlignedCodonFrame acf = new AlignedCodonFrame();
- MapList ml = new MapList(new int[] { 1, 12 }, new int[] { 1, 4 }, 3, 1);
- acf.addMap(al1.getSequenceAt(0), al2.getSequenceAt(0), ml);
- acf.addMap(al1.getSequenceAt(1), al2.getSequenceAt(1), ml);
- al2.addCodonFrame(acf);
+ makeMappings(al1, al2);
((Alignment) al2).alignAs(al1, false, true);
assertEquals("K-Q-Y-L-", al2.getSequenceAt(0).getSequenceAsString());
}
/**
+ * Aligning protein from cDNA for a single sequence. This is the 'simple' case
+ * (as there is no need to compute codon 'alignments') but worth testing
+ * before tackling the multiple sequence case.
+ *
+ * @throws IOException
+ */
+ @Test(groups = { "Functional" })
+ public void testAlignAs_proteinAsCdna_singleSequence() throws IOException
+ {
+ /*
+ * simplest case remove all gaps
+ */
+ verifyAlignAs(">protein\n-Q-K-\n", ">dna\nCAAaaa\n", "QK");
+
+ /*
+ * with sequence offsets
+ */
+ verifyAlignAs(">protein/12-13\n-Q-K-\n", ">dna/20-25\nCAAaaa\n", "QK");
+ }
+
+ /**
* Test aligning cdna as per protein alignment.
*
* @throws IOException
*/
AlignmentI al1 = loadAlignment(CDNA_SEQS_1, "FASTA");
AlignmentI al2 = loadAlignment(AA_SEQS_1, "FASTA");
- AlignedCodonFrame acf = new AlignedCodonFrame();
- MapList ml = new MapList(new int[] { 1, 12 }, new int[] { 1, 4 }, 3, 1);
- acf.addMap(al1.getSequenceAt(0), al2.getSequenceAt(0), ml);
- acf.addMap(al1.getSequenceAt(1), al2.getSequenceAt(1), ml);
- al2.addCodonFrame(acf);
+ makeMappings(al1, al2);
/*
* Realign DNA; currently keeping existing gaps in introns only
}
/**
+ * Test aligning cdna as per protein - single sequences
+ *
+ * @throws IOException
+ */
+ @Test(groups = { "Functional" })
+ public void testAlignAs_cdnaAsProtein_singleSequence() throws IOException
+ {
+ /*
+ * simple case insert one gap
+ */
+ verifyAlignAs(">dna\nCAAaaa\n", ">protein\nQ-K\n", "CAA---aaa");
+
+ /*
+ * simple case but with sequence offsets
+ */
+ verifyAlignAs(">dna/5-10\nCAAaaa\n", ">protein/20-21\nQ-K\n",
+ "CAA---aaa");
+
+ /*
+ * insert gaps as per protein, drop gaps within codons
+ */
+ verifyAlignAs(">dna/10-18\nCA-Aa-aa--AGA\n", ">aa/6-8\n-Q-K--R\n",
+ "---CAA---aaa------AGA");
+ }
+
+ /**
+ * Helper method that makes mappings and then aligns the first alignment as
+ * the second
+ *
+ * @param fromSeqs
+ * @param toSeqs
+ * @param expected
+ * @throws IOException
+ */
+ public void verifyAlignAs(String fromSeqs, String toSeqs, String expected)
+ throws IOException
+ {
+ /*
+ * Load alignments and add mappings from nucleotide to protein (or from
+ * first to second if both the same type)
+ */
+ AlignmentI al1 = loadAlignment(fromSeqs, "FASTA");
+ AlignmentI al2 = loadAlignment(toSeqs, "FASTA");
+ makeMappings(al1, al2);
+
+ /*
+ * Realign DNA; currently keeping existing gaps in introns only
+ */
+ ((Alignment) al1).alignAs(al2, false, true);
+ assertEquals(expected, al1.getSequenceAt(0).getSequenceAsString());
+ }
+
+ /**
+ * Helper method to make mappings from protein to dna sequences, and add the
+ * mappings to the protein alignment
+ *
+ * @param alFrom
+ * @param alTo
+ */
+ public void makeMappings(AlignmentI alFrom, AlignmentI alTo)
+ {
+ AlignmentI prot = !alFrom.isNucleotide() ? alFrom : alTo;
+ AlignmentI nuc = alFrom == prot ? alTo : alFrom;
+
+ int ratio = (alFrom.isNucleotide() == alTo.isNucleotide() ? 1 : 3);
+
+ AlignedCodonFrame acf = new AlignedCodonFrame();
+
+ for (int i = 0; i < nuc.getHeight(); i++)
+ {
+ SequenceI seqFrom = nuc.getSequenceAt(i);
+ SequenceI seqTo = prot.getSequenceAt(i);
+ MapList ml = new MapList(new int[] { seqFrom.getStart(),
+ seqFrom.getEnd() },
+ new int[] { seqTo.getStart(), seqTo.getEnd() }, ratio, 1);
+ acf.addMap(seqFrom, seqTo, ml);
+ }
+
+ prot.addCodonFrame(acf);
+ }
+
+ /**
* Test aligning dna as per protein alignment, for the case where there are
* introns (i.e. some dna sites have no mapping from a peptide).
*
*/
String dna1 = "A-Aa-gG-GCC-cT-TT";
String dna2 = "c--CCGgg-TT--T-AA-A";
- AlignmentI al1 = loadAlignment(">Seq1\n" + dna1 + "\n>Seq2\n" + dna2
- + "\n", "FASTA");
- AlignmentI al2 = loadAlignment(">Seq1\n-P--YK\n>Seq2\nG-T--F\n",
- "FASTA");
+ AlignmentI al1 = loadAlignment(">Seq1/6-17\n" + dna1
+ + "\n>Seq2/20-31\n" + dna2 + "\n", "FASTA");
+ AlignmentI al2 = loadAlignment(
+ ">Seq1/7-9\n-P--YK\n>Seq2/11-13\nG-T--F\n", "FASTA");
AlignedCodonFrame acf = new AlignedCodonFrame();
// Seq1 has intron at dna positions 3,4,9 so splice is AAG GCC TTT
// Seq2 has intron at dna positions 1,5,6 so splice is CCG TTT AAA
- MapList ml1 = new MapList(new int[] { 1, 2, 5, 8, 10, 12 }, new int[] {
- 1, 3 }, 3, 1);
+ // TODO sequence offsets
+ MapList ml1 = new MapList(new int[] { 6, 7, 10, 13, 15, 17 }, new int[]
+ { 7, 9 }, 3, 1);
acf.addMap(al1.getSequenceAt(0), al2.getSequenceAt(0), ml1);
- MapList ml2 = new MapList(new int[] { 2, 4, 7, 12 },
- new int[] { 1, 3 }, 3, 1);
+ MapList ml2 = new MapList(new int[] { 21, 23, 26, 31 }, new int[] { 11,
+ 13 }, 3, 1);
acf.addMap(al1.getSequenceAt(1), al2.getSequenceAt(1), ml2);
al2.addCodonFrame(acf);