[khmer] counting 31-mers with khmer

[Qingpeng Zhang]

Hi Rajat,

(It seems that I failed to send my last emails to the mail list so I
send this again.)


As in the script Titus wrote[1], you can still use method get() to
retrieve the count of a kmer(as in ktable)

ht = khmer.new_counting_hash(K, HT_SIZE, N_HT)
ht.consume_fasta(filename)
count = ht.get('AACT')

Also, if you want to get a list of all the kmers and their counts,
here is a script for your reference.

import khmer
# here k=5
def get_count(filename):

        ht = khmer.new_counting_hash(5, 1000000, 2)
        ht.consume_fasta(filename)

        for i in range(4**5):
                if ht.get(i):
                        kmer = khmer.reverse_hash(i,5)
                        count = ht.get(i)
                        print kmer,count

file = 'test.fa'
get_count(file)

Hope this helps.



[1]  http://lists.idyll.org/pipermail/khmer/2013-June/000104.html



Regards,
Qingpeng


On Thu, Jun 20, 2013 at 2:25 PM, Rajat <rajatshuvro at gmail.com> wrote:
> Thanks. I got it running with load-into-counting.py. However, I want to
> print out the k-mers with frequencies in a file. I don't see any example for
> doing that in http://khmer.readthedocs.org/en/latest/scripts.html. Can
> anyone help?
>
> Thanks in advance.
> Rajat
>
>
> On Thu, Jun 20, 2013 at 9:27 AM, C. Titus Brown <ctb at msu.edu> wrote:
>>
>> On Thu, Jun 20, 2013 at 08:31:52AM -0400, Rajat wrote:
>> > Hi Prof ,
>> > I tried to follow the example at
>> > http://khmer.readthedocs.org/en/latest/ktable.html to extract kmers
>> > using
>> > khmer. It works fine for small values of k (<=15) but exits without any
>> > output as soon as I make k>=16. In
>> > http://khmer.readthedocs.org/en/latest/ktable.html they mention that it
>> > does not work for k>12 on their machines. However, it is mentioned in
>> > http://khmer.readthedocs.org/en/latest/introduction.html that khmer can
>> > count kmers for k<=32. Could you point me to some examples that I can
>> > follow to count 31 mers with khmer?
>>
>> Hi Rajat,
>>
>> please see this recent thread on the khmer mailing list:
>>
>> http://lists.idyll.org/pipermail/khmer/2013-June/thread.html#start
>>
>> Briefly, ktable does the "dumb thing" and allocates 4**k bytes of memory,
>> which only works up to ~k=15 on many computers.  The counting hash data
>> structure will count for much larger k.
>>
>> You can also take a look at this IPython Notebook:
>>
>>
>> http://nbviewer.ipython.org/urls/raw.github.com/ngs-docs/ngs-notebooks/master/ngs-5x-kmer-abundance-distributions-2013.ipynb
>>
>> for information on getting k-mer abundance distributions.
>>
>> best,
>> --titus
>> --
>> C. Titus Brown, ctb at msu.edu
>
>
>
> _______________________________________________
> khmer mailing list
> khmer at lists.idyll.org
> http://lists.idyll.org/listinfo/khmer
>