[Denovoassembler-devel] RE : RE : Why flush both k-mer and its complement?

[Sébastien B. <seb...@ul...>]

No need to be sorry !

Basically, this code extracts the k-mers from reads and send them to MPI peers.

2 reverse-complement k-mers will always go to the same destination.


In fact, only the lowest k-mer needs to be appended as the other is not used anyway in 

MessageProcessor::call_RAY_MPI_TAG_KMER_ACADEMY_DATA (communication/MessageProcessor.cpp)



call_RAY_MPI_TAG_KMER_ACADEMY_DATA processes messages with tag RAY_MPI_TAG_KMER_ACADEMY_DATA


So basically, if your patch is modified to flush the lowest k-mer among a and b, then it should work fine.




> On 21/07/11 17:39, Sébastien Boisvert wrote:
>> I think there is a bug:
>>
>> In your for loop, the second iteration (flush=1) should add b in the m_bufferedData, not a twice !
> 
> Ah, sorry, it does actually add the raw bits of a/b as well. I just
> noticed that the rankToFlush would be the same, and didn't think about
> copying the actual sequence.
> 
> -- David

                                                     Sébastien