Plotting length distribution of contigs?

[kga1978]

Hi all,

I have a fasta file with contigs and I would like to plot the lengths of the sequences. I have only found prinseq-graphs that can do that, but it requires me to install a bunch of perl modules (I believe - I can't find any installation information). Is there an easier way of doing this? Preferably a simple script or java program. Here's an example of some of my data:

>NODE_4_length_492_cov_13.477642
ACGGAGTATGACTTTGTATTGGTGGGTCCTTGCACTGAACCAGCCCCTCTGGTTGTGCAT
AGGGGAGGCTTGTGGGAATGTGGAAAGAAATTGGCGTCCTTTACACCTGTTATACAAGAC
CAGGATCTTGAAGTATTTGTGAGAGAGGTTGGGGACACTTCGTCTGACCTGCTGATTGGG
GCATTGAGTGATATGATGATAGACAGGCTGGGGTTAAGGGTGCAGTGGTCAGGGGTGGAC
ATTGTCTCCACACTTAGGGCTGCAGCGCCGAACTGCGAGGGGATCTTGAGTGCGGTTCTT
GAGGCAGTGGACAACTGGGTGGAGTTCAAAGGTTATGCTCTCTGTTATAGTAAGTCAAAG
GGGAAGGTGATGGTGCAGTCAAGTGGTGGTAAATTGAGACTGAAGGGCAGAACATGTGAG
GAGTTGACTAGGAAGGATGAATGCATCGAAGACATTGAGTAGTCTCCTGGCGATGGTTGG
CTCCCCCGGGGGGGCCCCCGGCGGGGGGTCCCCC
>NODE_7_length_554_cov_17.906137
ATTTATTTTGAGTCTTATGTGAAACCACGTGAAGGACCCCAATGTTCTTGTAGTCGCAAC
AAATGGTCTCACATAAGACTCAAAATAAATCTGCCTCATGAAATTGTCAACAGCATCACT
AGTGCTCACCACTCTTTCCTCCACTATGGGTTCATGTGTCCTACTGTGAGACAGCCTCAA
TTCAGATGATAACACAATGTAATGTTCCTCTCTTTTCCATTTCACAATATGTGAGACAAG
AGATAAGGCTTCACAGTTAACATCCAACGCAACACAGAGATCTAGGAATTTTATTCTAGG
TGACCACTTCATTTTGGTTGACGCTAGATCACTCATGAATGGCAATATGTGCTTCTCAAA
CACCGATGGGTACAGCCTTCTCAAAGAATGAATGATGTGATTCAAACCAACCCTATCCTC
TAATAGTTTTGATGCAGTTGGCTTTAAAGGAAAATAGTCACAAGGGTTATGCTTGAAAAA
ATCCAATACCTTAACTGTCTTAGGTTCCCCTAAGACCCATGCACCCAACTCTATTGCAGT
TGATAAGGAGATGCACATATAATCCCATAACAAGGG
>NODE_8_length_274_cov_16.138685
CCAAAATAAGTTGTCTTCCACTTTCACTCGAGGTGCGCAGAAATTGCTATCTGAAGCTAT
CAACAAGTCTGCATTCCAGAGCTCCATTGCATCTGGCTTTGTGGGGTTATGCAGAACATT
GGGTAGCAAATGTGTTCGGGGACCAAATAAGGAGAATCTGTATATTAAGTCCATTCAGTC
TCTGATTTCTGATGTCAAGGGAATCAAATTATTGACAAATTCTAATGGCATTCAGTATTG
GCGGGTTCCGCTAGAACTTAGAGATGGGAGTGGAAGTGAAAGTGTGGTCAGTTATT

I would like to plot the lengths similar to this:

[maasha]

Using Biopieces you can use plot_distribution like this:

read_fasta -i contigs.fna | plot_distribution -k SEQ_LEN -x


                                  Distribution
      +          +          +          +           +          +          +
  14 ++*---------+----------+----------+-----------+----------+----------+-++
      |*                                                                   |
  12 ++*                                                                   ++
      |*                                                                   |
      |*                                                                   |
  10 ++*                                                                   ++
      |*                                                                   |
   8 ++*                                                                   ++
      |** *                                                                |
   6 ++****                                                                ++
      |****                                                                |
   4 ++*******                                                             ++
      |*******                                                             |
      |***********       * *                                               |
   2 ++**********************                                              ++
      |******************************************* **  * *            *    *
   0 ++*******************************************-**--*-*----+-------*--+-*+
      +          +          +          +           +          +          +
      0        10000      20000      30000       40000      50000      60000
                                     SEQ_LEN

It is also possible to output to X11 terminal or PNG, PDF, PS, and SVG.

Installing Biopieces requires a working setup of Perl w. relevant modules and Ruby w. relevant gems. However, it should be worth the trouble since Biopieces is a nice toolbox IMHO.

[maubp]

There's a "Histogram of sequence lengths" example in the Biopython Tutorial, it uses

• Python
• Biopython
• MatPlotLib (pylab)

Not Java, but also not Perl