NickAkhmetov/HMP-540/Update get_view_config_builder to use get_assaytype instead of get_assay

README.md

@@ -21,25 +21,25 @@ $ cd portal-visualization
 $ pip install .
 ...
 $ src/vis-preview.py --help
-usage: vis-preview.py [-h] (--url URL | --json JSON) [--types_url URL]
+usage: vis-preview.py [-h] (--url URL | --json JSON) [--assaytypes_url URL]
                       [--assets_url URL] [--token TOKEN] [--marker MARKER]
                       [--to_json]
 
 Given HuBMAP Dataset JSON, generate a Vitessce viewconf, and load vitessce.io.
 
 optional arguments:
-  -h, --help        show this help message and exit
-  --url URL         URL which returns Dataset JSON
-  --json JSON       File containing Dataset JSON
-  --types_url URL   Type service; default:
-                    https://search.api.hubmapconsortium.org/v3
-  --assets_url URL  Assets endpoint; default:
-                    https://assets.hubmapconsortium.org
-  --token TOKEN     Globus groups token; Only needed if data is not public
-  --marker MARKER   Marker to highlight in visualization; Only used in some
-                    visualizations.
-  --to_json         Output viewconf, rather than open in browser.
-```
+  -h, --help            show this help message and exit
+  --url URL             URL which returns Dataset JSON
+  --json JSON           File containing Dataset JSON
+  --assaytypes_url URL  AssayType service; default:
+                        https://ingest.api.hubmapconsortium.org/assaytype/
+  --assets_url URL      Assets endpoint; default:
+                        https://assets.hubmapconsortium.org
+  --token TOKEN         Globus groups token; Only needed if data is not public
+  --marker MARKER       Marker to highlight in visualization; Only used in
+                        some visualizations.
+  --to_json             Output viewconf, rather than open in browser.
+  ```
 
 ## Background
 

VERSION.txt

@@ -1 +1 @@
-0.0.13
+0.1.0

src/defaults.json

@@ -1,4 +1,4 @@
 {
-  "types_url": "https://search.api.hubmapconsortium.org/v3",
-  "assets_url": "https://assets.hubmapconsortium.org"
+  "assets_url": "https://assets.hubmapconsortium.org",
+  "assaytypes_url": "https://ingest.api.hubmapconsortium.org/assaytype/"
 }

src/portal_visualization/assays.py

@@ -11,3 +11,4 @@
 SCATAC_SEQ_SCI = "sc_atac_seq_sci"
 SCATAC_SEQ_SNARE = "sc_atac_seq_snare"
 SCATAC_SEQ_SN = "sn_atac_seq"
+SALMON_RNASSEQ_SLIDE = "salmon_rnaseq_slideseq"

src/portal_visualization/builder_factory.py

@@ -18,47 +18,76 @@
 from .assays import (
     SEQFISH,
     MALDI_IMS,
-    NANODESI
+    NANODESI,
+    SALMON_RNASSEQ_SLIDE
 )
 
 
-def get_view_config_builder(entity, get_assay):
-    data_types = entity["data_types"]
-    assay_objs = [get_assay(dt) for dt in data_types]
-    assay_names = [assay.name for assay in assay_objs]
-    hints = [hint for assay in assay_objs for hint in assay.vitessce_hints]
+def get_ancestor_assaytypes(entity, get_assaytype):
+    return [get_assaytype(ancestor).get('assaytype')
+            for ancestor
+            in entity.get('immediate_ancestors')]
+
+
+# This function is the main entrypoint for the builder factory.
+# It returns the correct builder for the given entity.
+#
+# The entity is a dict that contains the entity UUID and metadata.
+# `get_assaytype` is a function which takes an entity UUID and returns
+# a dict containing the assaytype and vitessce-hints for that entity.
+def get_view_config_builder(entity, get_assaytype):
+    if (entity.get('uuid') is None):
+        raise ValueError("Provided entity does not have a uuid")
+    assay = get_assaytype(entity)
+    assay_name = assay.get('assaytype')
+    hints = assay.get('vitessce-hints', [])
+
     dag_provenance_list = entity.get('metadata', {}).get('dag_provenance_list', [])
     dag_names = [dag['name']
                  for dag in dag_provenance_list if 'name' in dag]
     if "is_image" in hints:
         if 'sprm' in hints and 'anndata' in hints:
+            # e.g. CellDIVE [DeepCell + SPRM]
+            # sample entity: c3be5650e93907b68ddbdb22b948db32
             return MultiImageSPRMAnndataViewConfBuilder
         if "codex" in hints:
             if ('sprm-to-anndata.cwl' in dag_names):
+                # e.g. 'CODEX [Cytokit + SPRM]
+                # sample entity: 43213991a54ce196d406707ffe2e86bd
                 return StitchedCytokitSPRMViewConfBuilder
+            # Cannot find an example of this in the wild, every CODEX entity has the
+            # sprm-to-anndata.cwl in its dag_provenance_list
             return TiledSPRMViewConfBuilder
-        # Both SeqFISH and IMS were submitted very early on, before the
-        # special image pyramid datasets existed.  Their assay names should be in
-        # the `entity["data_types"]` while newer ones, like NanoDESI, are in the parents
-        if SEQFISH in assay_names:
+        # Check types of image pyramid ancestors to determine which builder to use
+        ancestor_assaytypes = get_ancestor_assaytypes(entity, get_assaytype)
+        # e.g. c6a254b2dc2ed46b002500ade163a7cc
+        if SEQFISH in [assaytype for assaytype in ancestor_assaytypes]:
             return SeqFISHViewConfBuilder
-        if MALDI_IMS in assay_names:
+        # e.g. 3bc3ad124014a632d558255626bf38c9
+        if MALDI_IMS in [assaytype for assaytype in ancestor_assaytypes]:
             return IMSViewConfBuilder
-        if NANODESI in [dt for e in entity["immediate_ancestors"] for dt in e["data_types"]]:
+        # e.g. 6b93107731199733f266bbd0f3bc9747
+        if NANODESI in [assaytype for assaytype in ancestor_assaytypes]:
             return NanoDESIViewConfBuilder
+        # e.g. f9ae931b8b49252f150d7f8bf1d2d13f
         return ImagePyramidViewConfBuilder
     if "rna" in hints:
         # This is the zarr-backed anndata pipeline.
         if "anndata-to-ui.cwl" in dag_names:
-            if "salmon_rnaseq_slideseq" in data_types:
+            if assay_name == SALMON_RNASSEQ_SLIDE:
+                # e.g. 2a590db3d7ab1e1512816b165d95cdcf
                 return SpatialRNASeqAnnDataZarrViewConfBuilder
+            # e.g. e65175561b4b17da5352e3837aa0e497
             return RNASeqAnnDataZarrViewConfBuilder
+        # e.g. c019a1cd35aab4d2b4a6ff221e92aaab
         return RNASeqViewConfBuilder
     if "atac" in hints:
+        # e.g. d4493657cde29702c5ed73932da5317c
         return ATACSeqViewConfBuilder
+    # any entity with no hints, e.g. 2c2179ea741d3bbb47772172a316a2bf
     return NullViewConfBuilder
 
 
-def has_visualization(entity, get_assay):
-    builder = get_view_config_builder(entity, get_assay)
+def has_visualization(entity, get_assaytype):
+    builder = get_view_config_builder(entity, get_assaytype)
     return builder != NullViewConfBuilder

src/vis-preview.py

@@ -9,15 +9,13 @@
 
 import requests
 
-from hubmap_commons.type_client import TypeClient
-
 from portal_visualization.builder_factory import get_view_config_builder
 
 
 def main():  # pragma: no cover
     defaults = json.load((Path(__file__).parent / 'defaults.json').open())
-    types_default_url = defaults['types_url']
     assets_default_url = defaults['assets_url']
+    assaytypes_default_url = defaults['assaytypes_url']
 
     parser = argparse.ArgumentParser(description='''
         Given HuBMAP Dataset JSON, generate a Vitessce viewconf, and load vitessce.io.''')
@@ -28,9 +26,9 @@ def main():  # pragma: no cover
         '--json', type=Path, help='File containing Dataset JSON')
 
     parser.add_argument(
-        '--types_url', metavar='URL',
-        help=f'Type service; default: {types_default_url}',
-        default=types_default_url)
+        '--assaytypes_url', metavar='URL',
+        help=f'AssayType service; default: {assaytypes_default_url}',
+        default=assaytypes_default_url)
     parser.add_argument(
         '--assets_url', metavar='URL',
         help=f'Assets endpoint; default: {assets_default_url}',
@@ -60,11 +58,13 @@ def main():  # pragma: no cover
         json_str = args.json.read_text()
     entity = json.loads(json_str)
 
-    def get_assay(name):
-        type_client = TypeClient(args.types_url)
-        return type_client.getAssayType(name)
+    def get_assaytype(uuid):
+        headers = {}
+        if args.token:
+            headers['Authorization'] = f'Bearer {args.token}'
+        requests.get(f'{defaults["assaytypes_url"]}/{uuid}', headers=headers).json()
 
-    Builder = get_view_config_builder(entity=entity, get_assay=get_assay)
+    Builder = get_view_config_builder(entity, get_assaytype)
     builder = Builder(entity, args.token, args.assets_url)
     print(f'Using: {builder.__class__.__name__}', file=stderr)
     conf_cells = builder.get_conf_cells(marker=marker)

test/assaytype-fixtures/04e7385339167e541ad42a2636e18398.json

@@ -0,0 +1,11 @@
+{
+    "assaytype": "codex_cytokit_v1",
+    "contains-pii": false,
+    "description": "CODEX [Cytokit + SPRM]",
+    "primary": false,
+    "vitessce-hints": [
+      "codex",
+      "is_image",
+      "is_tiled"
+    ]
+  }

test/assaytype-fixtures/210d118a14c8624b6bb9610a9062656e.json

@@ -0,0 +1,10 @@
+{
+    "assaytype": "sn_atac_seq",
+    "contains-pii": false,
+    "description": "snATAC-seq [SnapATAC]",
+    "primary": false,
+    "vitessce-hints": [
+      "is_sc",
+      "atac"
+    ]
+  }

test/assaytype-fixtures/2a590db3d7ab1e1512816b165d95cdcf.json

@@ -0,0 +1,10 @@
+{
+    "assaytype": "salmon_rnaseq_slideseq",
+    "contains-pii": false,
+    "description": "Slide-seq [Salmon]",
+    "primary": false,
+    "vitessce-hints": [
+      "is_sc",
+      "rna"
+    ]
+  }

test/assaytype-fixtures/2c2179ea741d3bbb47772172a316a2bf.json

@@ -0,0 +1,9 @@
+{
+    "assaytype": "bulk-RNA",
+    "contains-pii": true,
+    "description": "Bulk RNA-seq",
+    "dir-schema": "bulkrnaseq-v0",
+    "primary": true,
+    "tbl-schema": "bulkrnaseq-v0",
+    "vitessce-hints": []
+  }

test/assaytype-fixtures/3bc3ad124014a632d558255626bf38c9.json

@@ -0,0 +1,11 @@
+{
+    "assaytype": "MALDI-IMS",
+    "contains-pii": false,
+    "description": "MALDI IMS",
+    "dir-schema": "ims-v0",
+    "primary": true,
+    "tbl-schema": "ims-v2",
+    "vitessce-hints": [
+      "maldi"
+    ]
+  }

test/assaytype-fixtures/43213991a54ce196d406707ffe2e86bd.json

@@ -0,0 +1,11 @@
+{
+  "assaytype": "codex_cytokit_v1",
+  "contains-pii": false,
+  "description": "CODEX [Cytokit + SPRM]",
+  "primary": false,
+  "vitessce-hints": [
+    "codex",
+    "is_image",
+    "is_tiled"
+  ]
+}

test/assaytype-fixtures/6b93107731199733f266bbd0f3bc9747.json

@@ -0,0 +1,9 @@
+{
+  "assaytype": "NanoDESI",
+  "contains-pii": false,
+  "description": "NanoDESI",
+  "dir-schema": "ims-v0",
+  "primary": true,
+  "tbl-schema": "ims-v2",
+  "vitessce-hints": []
+}

test/assaytype-fixtures/9db61adfc017670a196ea9b3ca1852a0.json

@@ -0,0 +1,11 @@
+{
+    "assaytype": "image_pyramid",
+    "contains-pii": false,
+    "description": "Image Pyramid",
+    "primary": false,
+    "vitessce-hints": [
+      "is_image",
+      "pyramid",
+      "is_support"
+    ]
+  }

test/assaytype-fixtures/a6116772446f6d1c1f6b3d2e9735cfe0.json

@@ -0,0 +1,11 @@
+{
+    "assaytype": "image_pyramid",
+    "contains-pii": false,
+    "description": "Image Pyramid",
+    "primary": false,
+    "vitessce-hints": [
+      "is_image",
+      "pyramid",
+      "is_support"
+    ]
+  }

test/assaytype-fixtures/c019a1cd35aab4d2b4a6ff221e92aaab.json

@@ -0,0 +1,10 @@
+{
+    "assaytype": "salmon_sn_rnaseq_10x",
+    "contains-pii": false,
+    "description": "snRNA-seq [Salmon]",
+    "primary": false,
+    "vitessce-hints": [
+      "is_sc",
+      "rna"
+    ]
+  }

test/assaytype-fixtures/c3be5650e93907b68ddbdb22b948db32.json

@@ -0,0 +1,12 @@
+{
+    "assaytype": "celldive_deepcell",
+    "contains-pii": false,
+    "description": "CellDIVE [DeepCell + SPRM]",
+    "primary": false,
+    "vitessce-hints": [
+      "sprm",
+      "anndata",
+      "is_image",
+      "is_tiled"
+    ]
+  }

test/assaytype-fixtures/c6a254b2dc2ed46b002500ade163a7cc.json

@@ -0,0 +1,9 @@
+{
+    "assaytype": "seqFish",
+    "contains-pii": false,
+    "description": "seqFISH",
+    "dir-schema": "seqfish-v0",
+    "primary": true,
+    "tbl-schema": "seqfish-v0",
+    "vitessce-hints": []
+  }

test/assaytype-fixtures/d4493657cde29702c5ed73932da5317c.json

@@ -0,0 +1,10 @@
+{
+    "assaytype": "sc_atac_seq_sci",
+    "contains-pii": false,
+    "description": "sciATAC-seq [SnapATAC]",
+    "primary": false,
+    "vitessce-hints": [
+      "is_sc",
+      "atac"
+    ]
+  }

test/assaytype-fixtures/e1c4370da5523ab5c9be581d1d76ca20.json

@@ -0,0 +1,11 @@
+{
+    "assaytype": "image_pyramid",
+    "contains-pii": false,
+    "description": "Image Pyramid",
+    "primary": false,
+    "vitessce-hints": [
+      "is_image",
+      "pyramid",
+      "is_support"
+    ]
+  }

test/assaytype-fixtures/e65175561b4b17da5352e3837aa0e497.json

@@ -0,0 +1,10 @@
+{
+    "assaytype": "salmon_sn_rnaseq_10x",
+    "contains-pii": false,
+    "description": "snRNA-seq [Salmon]",
+    "primary": false,
+    "vitessce-hints": [
+      "is_sc",
+      "rna"
+    ]
+  }

test/assaytype-fixtures/ea4cfecb8495b36694d9a951510dc3c6.json

@@ -0,0 +1,10 @@
+{
+    "assaytype": "salmon_rnaseq_slideseq",
+    "contains-pii": false,
+    "description": "Slide-seq [Salmon]",
+    "primary": false,
+    "vitessce-hints": [
+      "is_sc",
+      "rna"
+    ]
+  }