[Biojava-dev] Bug in PHYLIPFileBuilder?

[Felipe Albrecht]

Hello, im trying to convert a protein multiple alignment in fasta
format to phylip format.

The source is:

		BufferedReader br = new BufferedReader(new FileReader(args[0]));
		PHYLIPFileBuilder builder = new PHYLIPFileBuilder();
		RichSequenceIterator richSequenceIterator =
IOTools.readFastaProtein(br, null);
		List<Sequence> l = new LinkedList<Sequence>();	
		Sequence seq = null;
		while (richSequenceIterator.hasNext()) {
			l.add(richSequenceIterator.nextSequence());		
		}
		builder.startFile();
		builder.setSequenceCount(l.size());
		builder.setSitesCount(seq.seqString().length());
		for (Sequence sequence : l) {
			builder.setCurrentSequenceName(sequence.getName());
			builder.receiveSequence(sequence.seqString());
		}
		builder.endFile();

As I said, my input data is a protein multiple alignment.

Running this source, this trace is showed:

Exception in thread "main" org.biojava.bio.BioError: Something has
gone badly wrong with DNA
	at org.biojava.bio.seq.DNATools.createDNASequence(DNATools.java:199)
	at org.biojava.bio.seq.DNATools.createGappedDNASequence(DNATools.java:207)
	at org.biojavax.bio.phylo.io.phylip.PHYLIPFileBuilder.createSequences(PHYLIPFileBuilder.java:121)
	at org.biojavax.bio.phylo.io.phylip.PHYLIPFileBuilder.buildAlignment(PHYLIPFileBuilder.java:94)
	at org.biojavax.bio.phylo.io.phylip.PHYLIPFileBuilder.endFile(PHYLIPFileBuilder.java:63)
	at FastaParser.main(FastaParser.java:54)
Caused by: org.biojava.bio.symbol.IllegalSymbolException: This
tokenization doesn't contain character: 'Q'
	at org.biojava.bio.seq.io.CharacterTokenization.parseTokenChar(CharacterTokenization.java:175)
	at org.biojava.bio.seq.io.CharacterTokenization$TPStreamParser.characters(CharacterTokenization.java:246)
	at org.biojava.bio.symbol.SimpleSymbolList.<init>(SimpleSymbolList.java:178)
	at org.biojava.bio.seq.DNATools.createDNA(DNATools.java:173)
	at org.biojava.bio.seq.DNATools.createDNASequence(DNATools.java:195)
	... 5 more

IMHO, the bug is at the line 121 of PHYLIPFileBuilder.java, method
createSequence , where is done:
        try {
          DNATools.createGappedDNASequence(sequence, name);
        } catch (IllegalSymbolException e) {
          isDNA = false;
        }

Where is execpeted that DNATools.CreateGappedDNASequence throws a
IllegalSymbolException , but seeking this method, in the file
DNAToos.java line 198:
    } catch (BioException se) {
      throw new BioError("Something has gone badly wrong with DNA", se);
    }
Being the IllegalSymbolException subclass of the BioError, they is
"catched" and a new exception is created and in
NATools.createSequences they arent catched.


I solved this problem adding:
    } catch (IllegalSymbolException ie) {
    	throw ie;
    }

in createDNASequence, but it's a workaround for the exception be catched.

A better solution, is check the type of the sequence. Exist a method
for discover is the sequence is DNA/RNA/Protein/mistake? If yes, uses
it, also the exceptions must be used when occurs an exception and dont
for flow control.

PS: Im using the biojava source code downloaded today from
http://www.biojava.org/download/bj15b/all/biojava-1.5-beta2.tar.gz

Thanks and Im waiting opinions.

Felipe Albrecht