Add flag for primary evidence to neo4j

scripts/cancer_cell_reproduction.py

@@ -75,6 +75,9 @@
     remove_imprecise=False,
     mutations=[
         Mutation(source="efo", confidence=0.7),
+        Mutation(source="bto", confidence=0.7),
+        Mutation(source="cl", confidence=0.7),
+        Mutation(source="clo", confidence=0.7),
         Mutation(source="depmap", confidence=0.7),
         Mutation(source="ccle", confidence=0.7),
         Mutation(source="cellosaurus", confidence=0.7),

src/semra/__init__.py

@@ -1,4 +1,4 @@
-from semra.rules import DB_XREF, EXACT_MATCH, LEXICAL_MAPPING, MANUAL_MAPPING, UNSPECIFIED_MAPPING, REPLACED_BY
+from semra.rules import DB_XREF, EXACT_MATCH, LEXICAL_MAPPING, MANUAL_MAPPING, REPLACED_BY, UNSPECIFIED_MAPPING
 from semra.struct import Evidence, Mapping, MappingSet, ReasonedEvidence, Reference, SimpleEvidence
 
 __all__ = [

src/semra/client.py

@@ -220,9 +220,9 @@ def get_highest_exact_matches(self, limit: int = 10) -> Counter:
         query = "MATCH (a)-[:`skos:exactMatch`]-(b) WHERE a.priority RETURN a.curie, count(distinct b) as c ORDER BY c DESCENDING LIMIT $limit"
         return Counter(dict(self.read_query(query, limit=limit)))
 
-    def get_exact_matches(self, curie: str) -> set[Reference]:
+    def get_exact_matches(self, curie: str) -> dict[Reference, str]:
         query = "MATCH (a {curie: $curie})-[:`skos:exactMatch`]-(b) RETURN b"
-        return {Reference.from_curie(node["curie"]) for node, in self.read_query(query, curie=curie)}
+        return {Reference.from_curie(node["curie"]): node["name"] for node, in self.read_query(query, curie=curie)}
 
     def get_concept_name(self, curie: str) -> str:
         return _get_name_by_curie(curie)

src/semra/io.py

@@ -5,7 +5,7 @@
 import pickle
 from pathlib import Path
 from textwrap import dedent
-from typing import TextIO, cast
+from typing import Literal, TextIO, cast
 
 import bioontologies
 import bioregistry
@@ -14,9 +14,9 @@
 import pandas as pd
 import pyobo
 import pyobo.utils
+from bioregistry import Collection
 from tqdm.auto import tqdm
 
-from bioregistry import Collection
 from semra.rules import DB_XREF, MANUAL_MAPPING, UNSPECIFIED_MAPPING
 from semra.struct import Evidence, Mapping, MappingSet, ReasonedEvidence, Reference, SimpleEvidence
 
@@ -411,10 +411,14 @@ def from_pickle(path: str | Path) -> list[Mapping]:
 
 
 def _edge_key(t):
-    s, p, o, c = t
+    s, p, o, c, *_ = t
     return s, p, o, 1 if isinstance(c, float) else 0, t
 
 
+def _neo4j_bool(b: bool, /) -> Literal["true", "false"]:  # noqa:FBT001
+    return "true" if b else "false"
+
+
 def write_neo4j(
     mappings: list[Mapping],
     directory: str | Path,
@@ -438,7 +442,7 @@ def write_neo4j(
         equivalence_classes = {}
 
     mapping_nodes_path = directory.joinpath("mapping_nodes.tsv")
-    mapping_nodes_header = ["curie:ID", ":LABEL", "prefix", "predicate", "confidence"]
+    mapping_nodes_header = ["curie:ID", ":LABEL", "prefix", "predicate", "confidence", "hasPrimary:boolean"]
 
     evidence_nodes_path = directory.joinpath("evidence_nodes.tsv")
     evidences = {}
@@ -453,11 +457,25 @@ def write_neo4j(
 
     mapping_set_nodes_path = directory.joinpath("mapping_set_nodes.tsv")
     mapping_sets = {}
-    mapping_set_nodes_header = ["curie:ID", ":LABEL", "prefix", "name", "license", "version", "confidence:float"]
+    mapping_set_nodes_header = [
+        "curie:ID",
+        ":LABEL",
+        "prefix",
+        "name",
+        "license",
+        "version",
+        "confidence:float",
+    ]
 
     edges_path = directory.joinpath("edges.tsv")
     edges: list[tuple[str, str, str, str | float]] = []
-    edges_header = [":START_ID", ":TYPE", ":END_ID", "confidence:float"]
+    edges_header = [
+        ":START_ID",
+        ":TYPE",
+        ":END_ID",
+        "confidence:float",
+        "hasPrimary:boolean",
+    ]
 
     for mapping in tqdm(mappings, unit="mapping", unit_scale=True, desc="Preparing Neo4j"):
         concepts.add(mapping.s)
@@ -468,6 +486,7 @@ def write_neo4j(
                 mapping.p.curie,
                 mapping.o.curie,
                 round(c, 4) if (c := mapping.confidence) is not None else "",
+                _neo4j_bool(mapping.has_primary_evidence),
             )
         )
         edges.append((mapping.curie, ANNOTATED_SOURCE.curie, mapping.s.curie, ""))
@@ -500,7 +519,7 @@ def write_neo4j(
                 "concept",
                 concept.prefix,
                 pyobo.get_name_by_curie(concept.curie) or "" if add_labels else "",
-                "true" if equivalence_classes.get(concept, False) else "false",
+                _neo4j_bool(equivalence_classes.get(concept, False)),
             )
             for concept in sorted(concepts, key=lambda n: n.curie)
         ),
@@ -509,7 +528,14 @@ def write_neo4j(
         mapping_nodes_path,
         mapping_nodes_header,
         (
-            (mapping.curie, "mapping", "semra.mapping", mapping.p.curie, mapping.confidence)
+            (
+                mapping.curie,
+                "mapping",
+                "semra.mapping",
+                mapping.p.curie,
+                mapping.confidence and round(mapping.confidence, 4),
+                _neo4j_bool(mapping.has_primary_evidence),
+            )
             for mapping in sorted(mappings, key=lambda n: n.curie)
         ),
     )
@@ -577,7 +603,7 @@ def write_neo4j(
             curl -sS https://bootstrap.pypa.io/get-pip.py | python3.11
 
         ARG twiddle1=dee
-        RUN python3.11 -m pip install git+https://github.com/biopragmatics/semra.git#egg=semra[web]
+        RUN python3.11 -m pip install "semra[web] @ git+https://github.com/biopragmatics/semra.git"
 
         # Add graph content
         ARG twiddle2=dee

src/semra/struct.py

@@ -225,6 +225,11 @@ def get_reference(self) -> Reference:
     def curie(self) -> str:
         return self.get_reference().curie
 
+    @property
+    def has_primary_evidence(self) -> bool:
+        """Get if there is a primary evidence associated with this mapping."""
+        return any(isinstance(evidence, SimpleEvidence) for evidence in self.evidence)
+
 
 def line(*references: Reference) -> list[Mapping]:
     """Create a list of mappings from a simple mappings path."""

src/semra/templates/concept.html

@@ -12,17 +12,20 @@
         <div class="card">
             <div class="card-body">
                 <h5 class="card-title">
-                    <a href="https://bioregistry.io/{{ curie }}">{{ curie }}</a>
+                    {{ name }} <a class="badge badge-info" href="https://bioregistry.io/{{ curie }}">{{ curie }}</a>
                 </h5>
                 <h6>Exact Matches</h6>
             </div>
             <table class="table table-striped table-borderless">
                 <tbody>
-                {% for exact_match in exact_matches | sort(attribute='curie') %}
+                {% for exact_match, name in exact_matches.items() %}
                 <tr>
                     <td>
                         <code>{{ exact_match.curie }}</code>
                     </td>
+                    <td>
+                        {{ name }}
+                    </td>
                     <td>
                         <a href="{{ url_for('view_concept', curie=exact_match.curie) }}">SeMRA</a>
                     </td>

src/semra/wsgi.py

@@ -2,8 +2,6 @@
 
 import os
 
-import biomappings.resources
-import biomappings.utils
 import fastapi
 import flask
 from curies import Reference
@@ -15,6 +13,11 @@
 from semra import Evidence, Mapping, MappingSet
 from semra.client import Neo4jClient
 
+try:
+    import biomappings.utils as biomappings_utils
+except ImportError:
+    biomappings_utils = None
+
 client = Neo4jClient()
 
 api_router = fastapi.APIRouter()
@@ -29,7 +32,7 @@
 api_router.mount("/", WSGIMiddleware(flask_app))
 
 EXAMPLE_MAPPINGS = ["25b67912bc720127a43a06ce4688b672", "5a56bf7ac409d8de84c3382a99e17715"]
-BIOMAPPINGS_GIT_HASH = biomappings.utils.get_git_hash()
+BIOMAPPINGS_GIT_HASH = biomappings_utils is not None and biomappings_utils.get_git_hash()
 
 PREDICATE_COUNTER = client.summarize_predicates()
 MAPPING_SET_COUNTER = client.summarize_mapping_sets()
@@ -93,11 +96,17 @@ def view_mapping(curie: str):
 def view_concept(curie: str):
     """View a concept."""
     reference = Reference.from_curie(curie)
+    name = client.get_concept_name(curie)
     exact_matches = client.get_exact_matches(curie)
     # TODO when showing equivalence between two entities from same namespace, suggest curating a replaced by relation
 
     return render_template(
-        "concept.html", reference=reference, curie=curie, exact_matches=exact_matches, has_biomappings=BIOMAPPINGS_GIT_HASH is not None
+        "concept.html",
+        reference=reference,
+        curie=curie,
+        name=name,
+        exact_matches=exact_matches,
+        has_biomappings=BIOMAPPINGS_GIT_HASH is not None,
     )
 
 
@@ -110,6 +119,8 @@ def mark_exact_incorrect(source: str, target: str):
         flask.flash("Can't interact with biomappings", category="error")
         return flask.redirect(flask.url_for(view_concept.__name__, curie=source))
 
+    import biomappings.resources
+
     source_reference = Reference.from_curie(source)
     target_reference = Reference.from_curie(target)