feat: add group-by-variant-oncoprint

.gitignore

@@ -6,7 +6,7 @@
 !workflow/scripts/*
 !workflow/rules/*
 !workflow/rules/annotation/*
-!workflow/resources/*
+!workflow/resources/**
 !config/*
 !LICENSE
 !README.md

config/config.yaml

@@ -67,6 +67,17 @@ report:
     activate: false
     # select a sample sheet column for stratification
     by-column: condition
+  # Optionally define annotation label types sort direction and domain for
+  # displaying them as heatmaps.
+  # By default, the annotated values are considered to be 'nominal',
+  # leading to categorical color coding without an order.
+  # Alternatively, you can set them as quantitative or ordinal here.
+  # The domain can take any values that are considered valid by vega-lite.
+  # annotation_labels:
+  #   mycolumn:
+  #     type: ordinal
+  #     ascending: true
+  #     domain: [-100, 0]
 
 calling:
   # Set to true to infer classical genotypes from Varlociraptor's VAF predictions
@@ -181,8 +192,6 @@ tables:
     # Uncomment and add VEP annotation fields to include (IMPACT, Consequence, Feature, SYMBOL, and HGVSp are always included)
     annotation_fields:
       - EXON
-    event_prob: true
-    observations: true
   generate_excel: true
 
 params:

workflow/envs/somalier.yaml

@@ -0,0 +1,7 @@
+channels:
+  - conda-forge
+  - bioconda
+  - nodefaults
+dependencies:
+  - somalier =0.2.18
+  - bcftools =1.17

workflow/resources/datavzrd/compact_oncoprint_header_labels.json

@@ -0,0 +1,6 @@
+[
+    "symbol",
+    "hgvsp",
+    "hgvsg",
+    "consequence"
+]

workflow/resources/datavzrd/spec_compact_oncoprint.json.j2

@@ -0,0 +1,180 @@
+{
+    "$schema": "https://vega.github.io/schema/vega-lite/v5.json",
+    "config": {
+        "concat": {
+            "spacing": 0
+        }
+    },
+    "resolve": {
+        "scale": {
+            "color": "independent",
+            "x": "shared"
+        }
+    },
+    "vconcat": [
+        {% for annotation_label in params.annotation_labels %}
+        {
+            "transform": [
+                {
+                    "filter": "datum.annotation == '{{ annotation_label|replace("'", "\'") }}'"
+                }
+            ],
+            "mark": "rect",
+            "encoding": {
+                "x": {
+                    "field": "group",
+                    "type": "ordinal",
+                    {% if loop.first %}
+                    "axis": {
+                        "orient": "top",
+                        "title": null
+                    },
+                    {% else %}
+                    "axis": null,
+                    {% endif %}
+                    "sort": {{ params.group_sorting|tojson }}
+                },
+                "y": {
+                    "field": "annotation",
+                    "axis": {
+                        "title": null
+                    },
+                    "scale": {
+                        "scheme": "category10"
+                    }
+                },
+                "color": {
+                    "field": "annotation_value",
+                    "type": {{ params.annotation_label_config[annotation_label].get("type", "nominal")|tojson }},
+                    {% if params.annotation_label_config[annotation_label].get("domain") is not none %}
+                    "scale": {
+                        "domain": {{ params.annotation_label_config[annotation_label].get("domain")|tojson }}
+                    },
+                    {% endif %}
+                    "legend": null
+                },
+                "opacity": {
+                    "value": 0.8
+                },
+                "tooltip": {"field": "annotation_value"}
+            }
+        },
+        {% endfor %}
+        {
+            "layer": [
+                {
+                    "transform": [
+                        {
+                            "filter": "datum.status"
+                        }
+                    ],
+                    "mark": {
+                        "type": "rect",
+                        "color": "lightgrey"
+                    },
+                    "encoding": {
+                        "x": {
+                            "field": "group",
+                            "type": "ordinal",
+                            "axis": null,
+                            "sort": {{ params.group_sorting|tojson }}
+                        },
+                        "y": {
+                            "field": "symbol",
+                            "type": "nominal",
+                            "axis": {
+                                "title": null
+                            },
+                            "sort": {
+                                "op": "count",
+                                "field": "status",
+                                "order": "descending"
+                            }
+                        }
+                    }
+                },
+                {
+                    "transform": [
+                        {
+                            "filter": "datum.status"
+                        }
+                    ],
+                    "height": {
+                        "step": 3
+                    },
+                    "mark": {
+                        "type": "rect",
+                        "tooltip": true,
+                    },
+                    "encoding": {
+                        "x": {
+                            "field": "group",
+                            "type": "ordinal",
+                            "axis": null,
+                            "sort": {{ params.group_sorting|tojson }}
+                        },
+                        "y": {
+                            "field": "symbol",
+                            "type": "nominal",
+                            "axis": {
+                                "title": null,
+                                "grid": true,
+                                "zindex": 0
+                            },
+                            "sort": {
+                                "op": "count",
+                                "field": "status",
+                                "order": "descending"
+                            }
+                        },
+                        "yOffset": {
+                            "field": "status",
+                            "type": "nominal"
+                        },
+                        "color": {
+                            "field": "status",
+                            "type": "nominal",
+                            "scale": {
+                                "scheme": "category10"
+                            },
+                            "legend": {"labelLimit": 1000}
+                        }
+                    }
+                }
+            ]
+        },
+        {
+            "transform": [{"filter": "datum.status"}],
+            "mark": "rect",
+            "encoding": {
+                "x": {
+                    "field": "group",
+                    "type": "ordinal",
+                    "axis": null,
+                    "sort": {{ params.group_sorting|tojson }}
+                },
+                "color": {
+                    "aggregate": "count",
+                    "field": "status",
+                    "legend": {
+                        "title": "#events"
+                    }
+                },
+                "tooltip": {
+                    "field": "status",
+                    "aggregate": "count"
+                }
+            }
+            }
+    ],
+    "transform": [
+        {
+            "calculate": "datum.hgvsp ? (datum.symbol + ':' + split(replace(datum.hgvsp, '%3D', '='), ':')[1]): datum.hgvsg",
+            "as": "variant"
+        },
+        {
+            "calculate": "datum.hgvsp ? datum.hgvsp : datum.hgvsg",
+            "as": "alteration"
+        }
+    ]
+}

workflow/resources/datavzrd/variant-calls-template.datavzrd.yaml

@@ -4,6 +4,7 @@ default-view: ?"overview" if input.variant_oncoprints else f"{params.groups[0]}-
 
 __definitions__:
   - import os
+  - import pandas as pd
   - |
     def read_file(path):
         return open(path, 'r').read()
@@ -30,6 +31,16 @@ __definitions__:
       }}
     }}
     """
+  - |
+    def get_annotation_scale(annotation):
+        # TODO remove this function once datavzrd does it internally (next release)
+        thetype = params.annotation_label_config[annotation].get("type", "ordinal")
+        if thetype == "quantitative":
+            return "linear"
+        elif thetype == "nominal":
+            return "ordinal"
+        else:
+            return thetype
 
 datasets:
   ?if input.variant_oncoprints:
@@ -41,9 +52,12 @@ datasets:
         gene details:
           column: symbol
           view: "overview-{value}"
+    compact-oncoprint:
+      path: ?params.compact_oncoprint
+      separator: "\t"
     ?for label in params.labels.index.values:
       ?label:
-        path: ?os.path.join(params.oncoprint_sorted_datasets, f"{label}.tsv")
+        path: ?os.path.join(input.compact_oncoprint_sortings, f"{label}.tsv")
         separator: "\t"
         headers: 2
     ?for gene, path in params.variant_oncoprints:
@@ -71,7 +85,7 @@ views:
   ?if params.gene_oncoprint:
     ?for view in ["gene-oncoprint"] + list(params.labels.index.values):
       __variables__:
-        labels: ?params.group_annotations.columns.values if view == "gene-oncoprint" else [view]
+        labels: ?list(params.group_annotations.columns.values) if view == "gene-oncoprint" else [view]
       ?"overview" if view == "gene-oncoprint" else f"by {view}":
         ?if view == "gene-oncoprint":
           desc: |
@@ -95,7 +109,7 @@ views:
                   label: ?annotation
                   plot:
                     heatmap:
-                      scale: ordinal
+                      scale: ?get_annotation_scale(annotation)
                       color-scheme: category20
           columns:
             symbol:
@@ -124,21 +138,28 @@ views:
                 ticks:
                   scale: linear
               optional: true
-
+
+    ?for sort_label, spec_path in zip(params.labels.index, input.specs_compact_oncoprint):
+      ?f"compact-oncoprint-by-{sort_label}":
+        desc: ?f"Compact oncoprint sorted by {sort_label}. Hover annotation rows with the mouse pointer to see values associated with colors."
+        dataset: compact-oncoprint
+        render-plot:
+          spec-path: ?spec_path
+
     ?for gene, path in params.variant_oncoprints:
       ?f"overview-{gene}":
         desc: ?f"Variants found in gene {gene}."
         dataset: ?f"variant-oncoprint-{gene}"
         hidden: true
         render-table:
-          ?if params.group_annotations.columns.values:
+          ?if list(params.group_annotations.columns.values):
             headers:
               ?for i, annotation in enumerate(params.group_annotations.columns.values):
                 ?i + 1:
                   label: ?annotation
                   plot:
                     heatmap:
-                      scale: ordinal
+                      scale: ?get_annotation_scale(annotation)
                       color-scheme: category20
           columns:
             consequence: