ValueError: The condensed distance matrix must contain only finite values.

[TSL-RamKrishna]

My command is:

average_nucleotide_identity.py -i tests/test_ani_data/ -o tests/test_ANIb_output \
-m ANIb -g --workers 4 --scheduler SLURM --gmethod seaborn

And I get the following error:

Traceback (most recent call last):
  File "/hpc-home/shrestha/.local/bin/average_nucleotide_identity.py", line 11, in <module>
    load_entry_point('pyani==0.3.0a0', 'console_scripts', 'average_nucleotide_identity.py')()
  File "/hpc-home/shrestha/.local/lib/python3.7/site-packages/pyani-0.3.0a0-py3.7.egg/pyani/scripts/average_nucleotide_identity.py", line 1019, in run_main
  File "/hpc-home/shrestha/.local/lib/python3.7/site-packages/pyani-0.3.0a0-py3.7.egg/pyani/scripts/average_nucleotide_identity.py", line 836, in draw
  File "/hpc-home/shrestha/.local/lib/python3.7/site-packages/pyani-0.3.0a0-py3.7.egg/pyani/pyani_graphics/sns/__init__.py", line 161, in heatmap
  File "/hpc-home/shrestha/.local/lib/python3.7/site-packages/pyani-0.3.0a0-py3.7.egg/pyani/pyani_graphics/sns/__init__.py", line 118, in get_clustermap
  File "/hpc-home/shrestha/miniconda3/lib/python3.7/site-packages/seaborn/_decorators.py", line 46, in inner_f
    return f(**kwargs)
  File "/hpc-home/shrestha/miniconda3/lib/python3.7/site-packages/seaborn/matrix.py", line 1406, in clustermap
    tree_kws=tree_kws, **kwargs)
  File "/hpc-home/shrestha/miniconda3/lib/python3.7/site-packages/seaborn/matrix.py", line 1222, in plot
    tree_kws=tree_kws)
  File "/hpc-home/shrestha/miniconda3/lib/python3.7/site-packages/seaborn/matrix.py", line 1068, in plot_dendrograms
    tree_kws=tree_kws
  File "/hpc-home/shrestha/miniconda3/lib/python3.7/site-packages/seaborn/_decorators.py", line 46, in inner_f
    return f(**kwargs)
  File "/hpc-home/shrestha/miniconda3/lib/python3.7/site-packages/seaborn/matrix.py", line 786, in dendrogram
    label=label, rotate=rotate)
  File "/hpc-home/shrestha/miniconda3/lib/python3.7/site-packages/seaborn/matrix.py", line 594, in __init__
    self.linkage = self.calculated_linkage
  File "/hpc-home/shrestha/miniconda3/lib/python3.7/site-packages/seaborn/matrix.py", line 661, in calculated_linkage
    return self._calculate_linkage_scipy()
  File "/hpc-home/shrestha/miniconda3/lib/python3.7/site-packages/seaborn/matrix.py", line 630, in _calculate_linkage_scipy
    metric=self.metric)
  File "/hpc-home/shrestha/miniconda3/lib/python3.7/site-packages/scipy/cluster/hierarchy.py", line 1057, in linkage
    raise ValueError("The condensed distance matrix must contain only "
ValueError: The condensed distance matrix must contain only finite values.

[TSL-RamKrishna]

After I got this error, I worked on the code and found that there was error running blastn commands.

Example of blastn command is:

blastn -out tests/test_ANIb_output/blastn_output/NC_011916_vs_NC_002696.blast_tab \
-query tests/test_ANIb_output/blastn_output/NC_011916-fragments.fna \
-db tests/test_ANIb_output/blastn_output/NC_002696.fna \
-xdrop_gap_final 150 \
-dust no \
-evalue 1e-15 \
-max_target_seqs 1 \
-outfmt '6 qseqid sseqid length mismatch pident nident qlen slen qstart qend sstart send positive ppos gaps' \
-task blastn

Though the blastn command are correct, it runs into the following error:

Error: Too many positional arguments (1), the offending value: qseqid

This is because of the single quote in the option -outfmt not be considered and so the qseqid becomes an argument for blastn which then reports as positional argument error.

The work around is, I saved the commands to a randomly generated filename and run the file as a bash script. Wolla, it runs successfully.

I have pushed my code updates to the forked version of pyani here:
https://github.com/TeamMacLean/pyani

[baileythegreen]

@widdowquinn What is to be done here? It seems like this is unlikely to be a general issue, but there's not much information to see what specific circumstances might lead to it.

[widdowquinn]

IIRC this was brought into the codebase at about the same time as the SLURM-related changes in #215 . Those changes caused a numberr of regressions and had issues such as hard-coded paths to TSL machines thay caused tests to fail and were tough to trace, so the ideas were reimplemented, rather than code merged.

[baileythegreen]

Okay, then we can probably just close it; unless you want to wait for a period of time first.