question about comparing two samples

[Lisa Wei]

Hello,

I was wondering how I would actually compare clonal architectures between, for example, primary and relapse tumours. Would I have to put in 2 separate input files (containing the SNVs and copy numbers) and run the PyClone analysis on these files together?

What would the command I have to run look like?

Thanks!

Lisa

[Andrew]

Hi Lisa,

You need to create two separate TSV files, one per sample. They can then be run together by setting up an appropriate YAML config. Simple is to use the run_analysis_pipeline command which just needs the TSV files and generates the YAML file for you. See details in the instructions on running https://bitbucket.org/aroth85/pyclone/wiki/Usage#markdown-header-tsv-input-file.

Cheers,

Andy

[Lisa Wei]

Thanks Andy! 

Also I'm running into another issue when I run the clustering command:

/home/lwei/miniconda2/bin/PyClone plot_clusters

I get the error:

Traceback (most recent call last):

  File "/home/lwei/miniconda2/bin/PyClone", line 11, in <module>

    load_entry_point('PyClone==0.13.1', 'console_scripts', 'PyClone')()

  File "/home/lwei/miniconda2/lib/python2.7/site-packages/pyclone/cli.py", line 78, in main

    args.func(args)

  File "/home/lwei/miniconda2/lib/python2.7/site-packages/pyclone/run.py", line 393, in cluster_plot

    thin=args.thin,

  File "/home/lwei/miniconda2/lib/python2.7/site-packages/pyclone/run.py", line 435, in _cluster_plot

    thin=thin

  File "/home/lwei/miniconda2/lib/python2.7/site-packages/pyclone/post_process/plot/clusters.py", line 254, in scatter_plot

    error_df=error_df

  File "/home/lwei/miniconda2/lib/python2.7/site-packages/pyclone/post_process/plot/_scatter.py", line 26, in plot_all_pairs

    fig = pp.figure(figsize=(size, size))

  File "/home/lwei/miniconda2/lib/python2.7/site-packages/matplotlib/pyplot.py", line 533, in figure

    **kwargs)

  File "/home/lwei/miniconda2/lib/python2.7/site-packages/matplotlib/backend_bases.py", line 161, in new_figure_manager

    return cls.new_figure_manager_given_figure(num, fig)

  File "/home/lwei/miniconda2/lib/python2.7/site-packages/matplotlib/backend_bases.py", line 167, in new_figure_manager_given_figure

    canvas = cls.FigureCanvas(figure)

  File "/home/lwei/miniconda2/lib/python2.7/site-packages/matplotlib/backends/backend_qt5agg.py", line 24, in __init__

    super(FigureCanvasQTAgg, self).__init__(figure=figure)

  File "/home/lwei/miniconda2/lib/python2.7/site-packages/matplotlib/backends/backend_qt5.py", line 234, in __init__

    _create_qApp()

  File "/home/lwei/miniconda2/lib/python2.7/site-packages/matplotlib/backends/backend_qt5.py", line 125, in _create_qApp

    raise RuntimeError('Invalid DISPLAY variable')

RuntimeError: Invalid DISPLAY variable

Do you know where this might be coming from?

Thanks!

Lisa

[Davide Cittaro]

When using paired samples (i.e. pre/post) the list of loci may change. I have indeed tsv (and then yaml) files with different mutations which, ideally, should be identified as mutations of a private clone which is appearing/disappearing. All mutations have been named with a common rule so that identical mutations have a common name. Unfortunately when analysis is performed I only get information about shared mutations, while private are apparently discarded. Am I missing something in the config file?

d

[Davide Cittaro]

Nevermind, I’ve just figured I need to add private variants in every file with zero counts, isn’t it?

[Andrew Roth]

Basically. You should ideally include the observed read counts i.e. ref/alt counts. You may have to manually extract these since your variant caller may not report the private mutations in other samples.

On Mon, May 6, 2019, 9:46 AM Davide Cittaro, <daweo...@gmail.com> wrote:

Nevermind, I’ve just figured I need to add private variants in every file with zero counts, isn’t it?

--
You received this message because you are subscribed to the Google Groups "Pyclone User Group" group.
To unsubscribe from this group and stop receiving emails from it, send an email to pyclone-user-gr...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/pyclone-user-group/719ef030-918b-4267-a7ef-31684425281d%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[Dalal Alsultan]

Hi Andy,

I am also comparing primary to met samples per patient. Has the link to instructions on how to create the YAML file changed? as I can't seem to locate it from here https://bitbucket.org/aroth85/pyclone/wiki/Usage#markdown-header-tsv-input-file

Thanks,

Dalal