Thanks!
Any advices on how to find the amount of abundance of a node that appeared in input sequences as kmer itself (not as its reverse complement)? Or do we lose this information?
Hi,
I've just started learning the basics of GATB. Not sure if it is meaningful, but I have a simple question:
Is it possible to force to create two separate nodes for a kmer and its reverse complement, instead of merging them into a single node with minimum value?
Shatlyk
Hello,
No, GATB does not allow to have two distinct nodes for a kmer and its reverse complement.
Thanks!
Any advices on how to find the amount of abundance of a node that appeared in input sequences as kmer itself (not as its reverse complement)? Or do we lose this information?
Yes, GATB loses this information currently. A kmer and its reverse complement are considered to be the same object.. If this is blocking for you, i.e. you would use GATB if it supported counting kmers and reverse complement separately, write me at rayan@gatb.fr
No, it does not block me now. Actually, I'm trying to grasp the basics of GATB, in order to figure out what should I add on top of GATB to use in my project. Thank you very much, for the help!
Ok, my answer was focused more on the de Bruijn graph than on the kmers abundances.
Strange enough, we never considered for GATB the possibility to count kmers as they appeared in input sequences, maybe because we were interested in creating de Bruijn graphs where a node holds both a kmer and its reverse complement.
However, I just tried some hack in the GATB code and it is indeed possible to count kmers the way you want. We will think about the way to cleanly include this hack in a future version of GATB because this may be interesting for other people as well.
Log in to answer this question.