Computing tri/di/mononucleotide context per region of the human genome

Question

[alhafidzhamdan]

I am looking for help to compute the nucleotide context of the human genome hg38, per region along the chromosomes. Preferably something like this.

GRanges object with 309581 ranges and 98 metadata columns:

       seqnames            ranges strand |       AAA       AAC       AAG       AAT       ACA       ACC       ACG       ACT
          <Rle>         <IRanges>  <Rle> | <numeric> <numeric> <numeric> <numeric> <numeric> <numeric> <numeric> <numeric>
   [1]        1           1-10000      + |         0         0         0         0         0         0         0         0
   [2]        1       10001-20000      + |    0.0104     0.015    0.0152    0.0056    0.0138    0.0224    0.0044    0.0106
   [3]        1       20001-30000      + |    0.0235    0.0127    0.0186    0.0154    0.0193    0.0155    0.0054    0.0147
   [4]        1       30001-40000      + |    0.0309    0.0104    0.0224    0.0174    0.0188    0.0129    0.0027    0.0127
   [5]        1       40001-50000      + |    0.0613    0.0217    0.0242    0.0347      0.03    0.0101    0.0025    0.0181

But for all tri/di/mononucleotide context. Any ideas? A

[Pierre Lindenbaum]

how did you get the output above ? what's wrong with this package ?

[alhafidzhamdan]

This is from the Fishhook package. Nothing wrong with it - it just does not show you how to generate nucleotide context for hg38.

http://mskilab.com/fishHook/tutorial.html

I've also asked the authors.

Answer 1

[JC]

Just search for kmer analysis tools, there are multiple packages/programs, like compseq in Emboss or time ago I wrote one in Perl,