Plotting

bycon Plot functionality

Starting with version v1.0.30 (2023-04-14) the bycon package added the ability to produce the typical Progenetix-style CNV histograms and CNV sample plots. These plots can both be generated by adding the respective output=... option to standard Beacon queries or through use of the bycon package to visualize data from local storage. Plotting from local .pgxseg ... file is possible but so far we don't provide a ready-made for that.

Changes 2023-10-23 The plotting of is now hadled by dedicated endpoints in /services:

• /services/collationplots for pre-computed frequency histograms and heatstrips
• /services/sampleplots for query-derived sample plots (including histograms)
  • plot selection through plotType= with options histoplot, samplesplot and histoheatplot
  • defaults to histoplot if unspecified

Plot types

CNV histograms - /services/collationplots

CNV histograms can be generated either (fast) for one or multiple of the "collations" i.e. samples sharing a common code (diagnosis, technnique...) or identifier (cell line id, PMID ...), or as single histogram for the output of a Beacon query.

A complete list of collations can be retrieved through the /services/collations/ endpoint, e.g. /services/collations?datasetIds=progenetix&output=text.

Please note that the datasetIds parameter will fall back to the default parameter if not indicated.

Examples

• /services/collationplots/?filters=NCIT:C35562,NCIT:C3709
  • a combination of 2 histograms
• /services/collationplots?filters=NCIT:C35562,NCIT:C3709&datasetIds=progenetix,cellz
  • a combination of 2 histograms
• /services/collationplots/?filters=pgx:icdom-85003,pgx:icdom-81703,pgx:icdom-87003,pgx:icdom-87203,pgx:icdom-94003,pgx:icdom-95003,pgx:icdom-81403&plot_title=CNV+Comparison&plot_area_height=50&plot_axis_y_max=80&plot_label_y_values=50
  • a collations based example showing the use of some extra parameters such as
    • plot_title
    • plot_area_height
    • plot_axis_y_max & plot_label_y_values

CNV sample plots - /services/sampleplots

Sample selection based plotting uses the standard bycon query stack for sample retrieval (i.e. aggregation over the data model) and then generates CNV plots from the found samples, either as clustered individual profiles or as binned frequency plot data (histograms or heatstrips).

CAVE: Sample plots are very time consuming due to the retrieval and plotting of all variants per sample.

Examples

• /beacon/sampleplots?filters=pgx:icdom-95003&plot_filter_empty_samples=y&plotGeneSymbols=MYCN&plotType=samplesplot&limit=200
  • this example is based on the histoplot example above, but based on individual sample retrieval and plotting and with some plot modifications:
    • limits the output to 100 samples (limit=100)
    • removes samples w/o CNVs (plot_filter_empty_samples=y)
• /beacon/sampleplots?filters=pgx:icdom-95003&plotGeneSymbols=MYCN&limit=100&plotType=samplesplot
  • this example gets samples for ICD-O Morphology 95003/3 (a.k.a. pgx:icdom-95003)
  • limits the output to the first 1000 samples (limit=1000)
  • adds a label for the MYCN gene
• /beacon/sampleplots?filters=pgx:icdom-95003&plotGeneSymbols=MYCN&limit=100
  • this is the same selection and labeling but defaulting to the histoplot option since no plotType parameter is indicated

Plot parameters

Plot parameters can be given both in snake_case and in the corresponding camelCase format (plot_area_width or plotAreaWidth).

Parameter definitions

Plot parameters