Matplotlib error by running summarize_taxa_through_plots.py

[salvatore lachina]

Hello,

I am new to this forum. I am using qiime on Debian 8 machine and when  run summarize_taxa_through_plots.py I take an error.

The command is:

 summarize_taxa_through_plots.py -i /.../observation_table_sorted.biom -o .../ko_summary_plots/ -f -s

and the output is:

 *** ERROR RAISED DURING STEP: Plot Taxonomy Summary
Command run was:
 plot_taxa_summary.py -i Scrivania/Salvo/shotgun/pear_assembly/reverse_blat_mapped/ko_summary_plots/observation_table_sorted_sorted_L2.txt,Scrivania/Salvo/shotgun/pear_assembly/reverse_blat_mapped/ko_summary_plots/observation_table_sorted_sorted_L3.txt,Scrivania/Salvo/shotgun/pear_assembly/reverse_blat_mapped/ko_summary_plots/observation_table_sorted_sorted_L4.txt,Scrivania/Salvo/shotgun/pear_assembly/reverse_blat_mapped/ko_summary_plots/observation_table_sorted_sorted_L5.txt,Scrivania/Salvo/shotgun/pear_assembly/reverse_blat_mapped/ko_summary_plots/observation_table_sorted_sorted_L6.txt -o Scrivania/Salvo/shotgun/pear_assembly/reverse_blat_mapped/ko_summary_plots//taxa_summary_plots/
Command returned exit status: 1
Stdout:

Stderr
/home/salvatore/miniconda3/envs/qiime2.7/lib/python2.7/site-packages/matplotlib/collections.py:590: FutureWarning: elementwise comparison failed; returning scalar instead, but in the future will perform elementwise comparison
  if self._edgecolors == str('face'):
Traceback (most recent call last):
  File "/home/salvatore/miniconda3/envs/qiime2.7/bin/plot_taxa_summary.py", line 4, in <module>
    __import__('pkg_resources').run_script('qiime==1.9.1', 'plot_taxa_summary.py')
  File "/home/salvatore/miniconda3/envs/qiime2.7/lib/python2.7/site-packages/setuptools-27.2.0-py2.7.egg/pkg_resources/__init__.py", line 744, in run_script
   
  File "/home/salvatore/miniconda3/envs/qiime2.7/lib/python2.7/site-packages/setuptools-27.2.0-py2.7.egg/pkg_resources/__init__.py", line 1499, in run_script
   
  File "/home/salvatore/miniconda3/envs/qiime2.7/lib/python2.7/site-packages/qiime-1.9.1-py2.7.egg-info/scripts/plot_taxa_summary.py", line 278, in <module>
    main()
  File "/home/salvatore/miniconda3/envs/qiime2.7/lib/python2.7/site-packages/qiime-1.9.1-py2.7.egg-info/scripts/plot_taxa_summary.py", line 274, in main
    resize_nth_label, label_type, include_html_legend)
  File "/home/salvatore/miniconda3/envs/qiime2.7/lib/python2.7/site-packages/qiime/plot_taxa_summary.py", line 1138, in make_all_charts
    resize_nth_label, label_type, include_html_legend))
  File "/home/salvatore/miniconda3/envs/qiime2.7/lib/python2.7/site-packages/qiime/plot_taxa_summary.py", line 1073, in get_counts
    label_type, include_html_legend))
  File "/home/salvatore/miniconda3/envs/qiime2.7/lib/python2.7/site-packages/qiime/plot_taxa_summary.py", line 861, in make_HTML_table
    props={'title': title})
  File "/home/salvatore/miniconda3/envs/qiime2.7/lib/python2.7/site-packages/qiime/plot_taxa_summary.py", line 615, in make_area_bar_chart
    savefig(img_abs, dpi=80, facecolor=background_color)
  File "/home/salvatore/miniconda3/envs/qiime2.7/lib/python2.7/site-packages/matplotlib/pyplot.py", line 577, in savefig
    res = fig.savefig(*args, **kwargs)
  File "/home/salvatore/miniconda3/envs/qiime2.7/lib/python2.7/site-packages/matplotlib/figure.py", line 1476, in savefig
    self.canvas.print_figure(*args, **kwargs)
  File "/home/salvatore/miniconda3/envs/qiime2.7/lib/python2.7/site-packages/matplotlib/backend_bases.py", line 2211, in print_figure
    **kwargs)
  File "/home/salvatore/miniconda3/envs/qiime2.7/lib/python2.7/site-packages/matplotlib/backends/backend_agg.py", line 521, in print_png
    FigureCanvasAgg.draw(self)
  File "/home/salvatore/miniconda3/envs/qiime2.7/lib/python2.7/site-packages/matplotlib/backends/backend_agg.py", line 469, in draw
    self.figure.draw(self.renderer)
  File "/home/salvatore/miniconda3/envs/qiime2.7/lib/python2.7/site-packages/matplotlib/artist.py", line 59, in draw_wrapper
    draw(artist, renderer, *args, **kwargs)
  File "/home/salvatore/miniconda3/envs/qiime2.7/lib/python2.7/site-packages/matplotlib/figure.py", line 1085, in draw
    func(*args)
  File "/home/salvatore/miniconda3/envs/qiime2.7/lib/python2.7/site-packages/matplotlib/artist.py", line 59, in draw_wrapper
    draw(artist, renderer, *args, **kwargs)
  File "/home/salvatore/miniconda3/envs/qiime2.7/lib/python2.7/site-packages/matplotlib/axes/_base.py", line 2110, in draw
    a.draw(renderer)
  File "/home/salvatore/miniconda3/envs/qiime2.7/lib/python2.7/site-packages/matplotlib/artist.py", line 59, in draw_wrapper
    draw(artist, renderer, *args, **kwargs)
  File "/home/salvatore/miniconda3/envs/qiime2.7/lib/python2.7/site-packages/matplotlib/collections.py", line 772, in draw
    Collection.draw(self, renderer)
  File "/home/salvatore/miniconda3/envs/qiime2.7/lib/python2.7/site-packages/matplotlib/artist.py", line 59, in draw_wrapper
    draw(artist, renderer, *args, **kwargs)
  File "/home/salvatore/miniconda3/envs/qiime2.7/lib/python2.7/site-packages/matplotlib/collections.py", line 312, in draw
    mpath.Path(offsets), transOffset, tuple(facecolors[0]))
  File "/home/salvatore/miniconda3/envs/qiime2.7/lib/python2.7/site-packages/matplotlib/backends/backend_agg.py", line 124, in draw_markers
    return self._renderer.draw_markers(*kl, **kw)
OverflowError: Allocated too many blocks


I tryed to set the file miniconda3/envs/qiime1/lib/python2.7/site-packages/matplotlib/mpl-data/matplotlibrc and set the backend from Qt4Agg to agg but it return the same error.

Do you have any ideas?

Thank you in advance.

Best regards

Salvatore

[justink]

Sounds like maybe just too large a task for your machine. Try filtering your observation table (with filter_otus_from_otu_table.py or some other way) until it's fairly small in samples and observations. If that works, then it might just be a data too large issue.

[salvatore lachina]

Hi Justin,

thank you for reply. I followed your advice but I take another type of error. This is the output after ran summarize_taxa_through_plots.

py.

What do you think about this?

*** ERROR RAISED DURING STEP: Plot Taxonomy Summary
Command run was:

 plot_taxa_summary.py -i Scrivania/Salvo/shotgun/pear_assembly/reverse_blat_mapped/ko_summary_plots/observation_table_sorted_hight_confidence_sorted_L2.txt,Scrivania/Salvo/shotgun/pear_assembly/reverse_blat_mapped/ko_summary_plots/observation_table_sorted_hight_confidence_sorted_L3.txt,Scrivania/Salvo/shotgun/pear_assembly/reverse_blat_mapped/ko_summary_plots/observation_table_sorted_hight_confidence_sorted_L4.txt,Scrivania/Salvo/shotgun/pear_assembly/reverse_blat_mapped/ko_summary_plots/observation_table_sorted_hight_confidence_sorted_L5.txt,Scrivania/Salvo/shotgun/pear_assembly/reverse_blat_mapped/ko_summary_plots/observation_table_sorted_hight_confidence_sorted_L6.txt -o Scrivania/Salvo/shotgun/pear_assembly/reverse_blat_mapped/ko_summary_plots//taxa_summary_plots/

Command returned exit status: 1
Stdout:

Stderr

Traceback (most recent call last):
  File "/home/salvatore/miniconda3/envs/qiime2.7/bin/plot_taxa_summary.py", line 4, in <module>
    __import__('pkg_resources').run_script('qiime==1.9.1', 'plot_taxa_summary.py')
  File "/home/salvatore/miniconda3/envs/qiime2.7/lib/python2.7/site-packages/setuptools-27.2.0-py2.7.egg/pkg_resources/__init__.py", line 744, in run_script
   
  File "/home/salvatore/miniconda3/envs/qiime2.7/lib/python2.7/site-packages/setuptools-27.2.0-py2.7.egg/pkg_resources/__init__.py", line 1499, in run_script
   
  File "/home/salvatore/miniconda3/envs/qiime2.7/lib/python2.7/site-packages/qiime-1.9.1-py2.7.egg-info/scripts/plot_taxa_summary.py", line 278, in <module>
    main()
  File "/home/salvatore/miniconda3/envs/qiime2.7/lib/python2.7/site-packages/qiime-1.9.1-py2.7.egg-info/scripts/plot_taxa_summary.py", line 274, in main
    resize_nth_label, label_type, include_html_legend)
  File "/home/salvatore/miniconda3/envs/qiime2.7/lib/python2.7/site-packages/qiime/plot_taxa_summary.py", line 1138, in make_all_charts
    resize_nth_label, label_type, include_html_legend))

  File "/home/salvatore/miniconda3/envs/qiime2.7/lib/python2.7/site-packages/qiime/plot_taxa_summary.py", line 984, in get_counts
    chart_type, False)
  File "/home/salvatore/miniconda3/envs/qiime2.7/lib/python2.7/site-packages/qiime/plot_taxa_summary.py", line 715, in get_fracs
    frac = float(n) / total
ZeroDivisionError: float division by zero

Thank you in advance.

Salvatore

[justink]

Hmm, now I think it's an issue with non-integer OTU table counts. Would you post the output of 'biom summarize-table -i observation_table_sorted.biom' ?

[salvatore lachina]

Hi Justin,

this is the output from biom summarize_table.

What do you think?

Best,

Salvatore

[justink]

Well, that says you have about 40 000 samples, most of which have zero sequences. So I think that's the problem. I think you've basically got a transposed table—your 'samples' are actually individual sequences.

I think go back to the earlier steps and make sure at each step you have the correct number of samples. Maybe give tutorial data a try as well to make sure that works fine.

[salvatore lachina]

Hi Justin,

this analysis is a shotgun metagenome on a single sample. To do this I followed this tutorial "http://qiime.org/tutorials/shotgun_analysis.html"

 after assembled my reads by using PEAR.

In detail I runned this commands:

run_pear.pl 

readFilter.pl

run_fastq_to_fasta.pl

parallel_map_reads_to_reference.py

sort_otu_table.py

summarize_taxa_through_plots.py

The first three scripts came from the microbiome_helper repository on github by Morgan Langill.

I am very interested to perform a functional analysis by using KO or other functional database (do you have any suggestion about database or a pipeline?)

What do you think about this pipeline?

Any suggestion?

Thank you very much for your help

Best,

Salvatore

[justink]

Salvatore, sorry I missed your reply.

I think there'll be a bunch of strange effects because you've got only one sample. Area plots for example draw lines between samples. I'm not sure qiime is really the best tool here.

But, I think if you're trying to get e.g. a single bar representing taxonomy for your one sample, that should work. Take a look at the input to parallel_map_reads_to_reference.py—I think that's in a bad format. The output of parallel_map_reads_to_reference.py should be a biom table with exactly one sample.