Hi Nick,
I feel I should come in on this issue as I was the original author. The code
was copied from another location [1] and seemed to work at the time (and a
*lot* faster than using a map). I do apologise for the bug but it's a simple
fix. I populate a case insensitive Codon map during the class' construction.
The code you're referring to does:
int arrayIndex = triplet.intValue();
//So long as we're within range then access
if(arrayIndex > -1 && arrayIndex < codonArray.length) {
target = codonArray[arrayIndex];
if (target != null) {
aminoAcid = target.getAminoAcid();
}
}
//Otherwise we have to use the Map
else {
target = quickLookup.get(triplet);
aminoAcid = target.getAminoAcid();
}
if(aminoAcid == null && translateNCodons()) {
aminoAcid = unknownAminoAcidCompound;
}
Reduce this to:
target = quickLookup.get(triplet);
aminoAcid = target.getAminoAcid();
if(aminoAcid == null && translateNCodons()) {
aminoAcid = unknownAminoAcidCompound;
}
That way we never use the array for lookups. Using the array based system is
faster but if it isn't safe then it should be removed.
Can someone with write access do the following:
* Add this hash collision error in as a test case in
https://github.com/sbliven/biojava/blob/master/biojava3-core/src/test/java/org/biojava3/core/sequence/TranslationTest.java
* Confirm the bug
* Remove all mention of the codonArray
* Confirm the bug's removal & commit/push
Thanks & once again sorry
Andy
[1] the exact place does escape me but something is telling me it was an emboss
package
On 20 Sep 2013, at 13:45, Nick England <[email protected]> wrote:
> Hara,
>
> Hmm this is rather odd. I get the same issue with that sequence with a
> custom engine as well.
>
> My code has:
> Builder builder = new TranscriptionEngine.Builder();
> builder.initMet(false);
> builder.translateNCodons(true);
> builder.trimStop(false);
> TranscriptionEngine engine = builder.build();
> Sequence<AminoAcidCompound> seq=engine.translate(new
> DNASequence("GTNTGTTAGTGT"));
> assertEquals("XC*C", seq.toString());
> Sequence<AminoAcidCompound> seq2=engine.translate(new
> DNASequence("ANAANG"));
> System.out.println(seq2);
> the first sequence translates as expected, but your sequence is translating
> as HR, when it should be XX. This looks like a pretty bad bug!
>
> Nick
>
>
> On 19 September 2013 19:59, Hara Dilley <[email protected]> wrote:
>
>> Hi,
>>
>> Is there an issue with the DNA Translation in biojava3.core?
>> It appears that it wants to translate "N" in certain cases
>> Executing:
>> new
>> DNASequence("ANAANG").getRNASequence().getProteinSequence().getSequenceAsString();
>> will produce aa HR.
>>
>> thanks
>> Hara
>>
>> ________________________________
>>
>> This email and any attachments thereto may contain private, confidential,
>> and privileged material for the sole use of the intended recipient. Any
>> review, copying, or distribution of this email (or any attachments thereto)
>> by others is strictly prohibited. If you are not the intended recipient,
>> please contact the sender immediately and permanently delete the original
>> and any copies of this email and any attachments thereto.
>>
>> _______________________________________________
>> Biojava-l mailing list - [email protected]
>> http://lists.open-bio.org/mailman/listinfo/biojava-l
>>
> _______________________________________________
> Biojava-l mailing list - [email protected]
> http://lists.open-bio.org/mailman/listinfo/biojava-l
_______________________________________________
Biojava-l mailing list - [email protected]
http://lists.open-bio.org/mailman/listinfo/biojava-l
