TransgeneExpressionHere: restore Expressed_in column

v2 prod's legacy ExpressionOverlapsHere result table carried an
Expressed_in column listing the anatomy class(es) each EP overlaps
(e.g. "pacemaker neuron"). The v1.10.x migration to VFBquery dropped
it; the TODO at the top of get_transgene_expression_here flagged the
missing column for a follow-up. This is that follow-up.

Cypher:
  - WITH now collects `anat.label + '----' + anat.short_form` per
    distinct anat per ep (within the Owlery closure from v1.14.0),
    alongside the existing pub_shorts collection.
  - RETURN projects `expressed_in` capped at the first 5 entries
    with a `+N more` suffix so wide closures stay readable. Items
    joined with `; ` (matching the Reference column's
    stringDelimiter convention) and entity ids wrapped with `----`
    (matching the entityDelimiter convention) so the v2 frontend
    component can extract the click target.

Schema:
  - AnatomyExpressedIn_to_schema gains `expressed_in` in
    preview_columns (between name and pubs).
  - get_transgene_expression_here's formatted-dict headers add
    `expressed_in: {title: "Expressed_in", type: "markdown",
    order: 1}`, shifting pubs/tags/template/technique orders one slot
    later.
  - encode_markdown_links is NOT called on expressed_in: the
    `; `-joined wire format is already what the V2
    QueryLinkArrayComponent expects to receive verbatim; running it
    through the markdown-link encoder would corrupt the
    `----`-wrapped entity ids.

Companion geppetto-vfb change switches the queryBuilderConfiguration
entry for `expressed_in` from QueryLinkComponent (single link) to
QueryLinkArrayComponent (clickable chip list), adds
stringDelimiter ";", and keeps the existing entityDelimiter ("----"),
entityIndex (1), and click action (window.addVfbId('$entity$')).

Verified locally against pdb for pacemaker neuron (FBbt_00006048):
  P{GAL4-per.BS}     -> "pacemaker neuron----FBbt_00006048"
  P{GAL4-tim.E}      -> "pacemaker neuron----FBbt_00006048"
  P{GSV6}GS10340     -> "pacemaker neuron----FBbt_00006048"
  P{cry-GAL4.E}      -> "pacemaker neuron----FBbt_00006048"
matching v2 prod's single-value Expressed_in display for the same
entity.

Patch bump 1.14.2 -> 1.14.3. New column triggers natural cache
invalidation via the result shape change.

Author: Robbie1977

setup.py

@@ -3,7 +3,7 @@
 
 here = path.abspath(path.dirname(__file__))
 
-__version__ = "1.14.2"
+__version__ = "1.14.3"
 
 # Get the long description from the README file
 with open(path.join(here, 'README.md')) as f:

src/vfbquery/vfb_queries.py

@@ -1960,7 +1960,7 @@ def TransgeneExpressionHere_to_schema(name, take_default):
     
     Query chain: Multi-step Owlery and Neo4j queries
     """
-    return Query(query="TransgeneExpressionHere", label=f"Transgene expression in {name}", function="get_transgene_expression_here", takes={"short_form": {"$and": ["Class", "Nervous_system", "Anatomy"]}, "default": take_default}, preview=5, preview_columns=["id", "name", "pubs", "tags", "template", "technique", "thumbnail"])
+    return Query(query="TransgeneExpressionHere", label=f"Transgene expression in {name}", function="get_transgene_expression_here", takes={"short_form": {"$and": ["Class", "Nervous_system", "Anatomy"]}, "default": take_default}, preview=5, preview_columns=["id", "name", "expressed_in", "pubs", "tags", "template", "technique", "thumbnail"])
 
 
 def FindStocks_to_schema(name, take_default):
@@ -5424,11 +5424,14 @@ def get_transgene_expression_here(anatomy_short_form: str, return_dataframe=True
     INSTANCEOF leaf subclasses, not the parent. Same closure pattern as
     get_instances (v1.12.8).
 
-    TODO: Expressed_in column. Prod surfaces it from the
-    anatomy_channel_image[].anatomy.label list — one chip per
-    representative image. Needs a small design decision on how to
-    render multiple values in a flat string column; deferred to a
-    follow-up so the rest of the columns can ship now.
+    v1.14.3: adds Expressed_in column. Each EP row carries the leaf
+    anatomy classes its Individuals are INSTANCEOF (within the Owlery
+    closure of the queried class), formatted as
+    `label----short_form; label----short_form; ...` so the geppetto-vfb
+    QueryLinkArrayComponent renders one clickable chip per anat
+    (calling window.addVfbId on click). Capped at the first 5 with
+    `+N more` overflow so EPs that overlap large closures stay
+    readable.
     """
     # Resolve the full subclass closure of the input anatomy class via
     # Owlery. Owlery handles OWL inference (equivalence classes, defined
@@ -5468,7 +5471,9 @@ def get_transgene_expression_here(anatomy_short_form: str, return_dataframe=True
     main_query = f"""
         MATCH (ep:Class:Expression_pattern)<-[ar:overlaps|part_of]-(:Individual)-[:INSTANCEOF]->(anat:Class)
         WHERE anat.short_form IN {anat_short_forms!r}
-        WITH ep, collect(DISTINCT ar.pub[0]) AS pub_shorts
+        WITH ep,
+             collect(DISTINCT ar.pub[0]) AS pub_shorts,
+             collect(DISTINCT anat.label + '----' + anat.short_form) AS anat_pairs
         ORDER BY ep.label
         {limit_clause}
         CALL {{
@@ -5496,6 +5501,11 @@ def get_transgene_expression_here(anatomy_short_form: str, return_dataframe=True
         RETURN
             ep.short_form AS id,
             apoc.text.format("[%s](%s)", [ep.label, ep.short_form]) AS name,
+            CASE
+                WHEN size(anat_pairs) <= 5
+                THEN apoc.text.join(anat_pairs, '; ')
+                ELSE apoc.text.join(anat_pairs[0..5], '; ') + '; +' + toString(size(anat_pairs)-5) + ' more'
+            END AS expressed_in,
             apoc.text.join(coalesce(ep.uniqueFacets, []), '|') AS tags,
             pubs,
             REPLACE(apoc.text.format("[%s](%s)", [COALESCE(templ.symbol[0], templ.label), templ.short_form]), '[null](null)', '') AS template,
@@ -5512,16 +5522,17 @@ def get_transgene_expression_here(anatomy_short_form: str, return_dataframe=True
         return df
     return {
         "headers": {
-            "id": {"title": "ID", "type": "selection_id", "order": -1},
-            "name": {"title": "Expression Pattern", "type": "markdown", "order": 0},
-            "pubs": {"title": "Publications", "type": "metadata", "order": 1},
-            "tags": {"title": "Tags", "type": "tags", "order": 2},
-            "template": {"title": "Template", "type": "markdown", "order": 3},
-            "technique": {"title": "Imaging Technique", "type": "text", "order": 4},
-            "thumbnail": {"title": "Thumbnail", "type": "markdown", "order": 9},
+            "id":           {"title": "ID",                "type": "selection_id", "order": -1},
+            "name":         {"title": "Expression Pattern","type": "markdown",     "order": 0},
+            "expressed_in": {"title": "Expressed_in",      "type": "markdown",     "order": 1},
+            "pubs":         {"title": "Publications",      "type": "metadata",     "order": 2},
+            "tags":         {"title": "Tags",              "type": "tags",         "order": 3},
+            "template":     {"title": "Template",          "type": "markdown",     "order": 4},
+            "technique":    {"title": "Imaging Technique", "type": "text",         "order": 5},
+            "thumbnail":    {"title": "Thumbnail",         "type": "markdown",     "order": 9},
         },
         "rows": [
-            {k: row[k] for k in ['id', 'name', 'pubs', 'tags', 'template', 'technique', 'thumbnail']}
+            {k: row[k] for k in ['id', 'name', 'expressed_in', 'pubs', 'tags', 'template', 'technique', 'thumbnail']}
             for row in safe_to_dict(df, sort_by_id=False)
         ],
         "count": total_count,