Expanding hints scope

src/model/hint.js

@@ -1,9 +1,18 @@
 import _ from 'lodash';
+import { ENTITY_TYPE } from './element/entity-type.js';
 
 // Define constants for Hint types, combining ORGANISM with ENTITY_TYPE
-const HINT_TYPE = Object.freeze({
-  ORGANISM: 'organism',
-});
+const HINT_TYPE = Object.freeze(
+  _.assign(
+    {
+      ORGANISM: 'organism',
+      DISEASE: 'disease',
+      CELL_LINE: 'cell_line',
+      VARIANT: 'variant',
+    },
+    ENTITY_TYPE,
+  ),
+);
 
 // Flatten the HINT_TYPE object to create an array of all hint types
 const HINT_TYPES = _.flatMap(HINT_TYPE);

src/server/routes/api/document/hint/pubtator.js

@@ -12,18 +12,32 @@ function map(bioCDocument) {
 
   // See Table 1 https://www.ncbi.nlm.nih.gov/research/pubtator3/tutorial
   const PUBTATOR_ANNOTATION_TYPE = Object.freeze({
+    GENE: 'Gene',
     SPECIES: 'Species',
-    // could add more types here when scope expands
+    CHEMICAL: 'Chemical',
+    DISEASE: 'Disease',
+    CELL_LINE: 'CellLine',
+    // VARIANT: 'Variant',
   });
   const PUBTATOR_DATABASE = Object.freeze({
+    ncbi_gene: 'ncbi_gene',
     ncbi_taxonomy: 'ncbi_taxonomy',
-    // could add more databases here when scope expands
+    ncbi_mesh: 'ncbi_mesh',
+    cvcl: 'cvcl',
+    litvar: 'litvar',
   });
   const entityTypes = new Map([
+    [PUBTATOR_ANNOTATION_TYPE.GENE, HINT_TYPE.GGP],
     [PUBTATOR_ANNOTATION_TYPE.SPECIES, HINT_TYPE.ORGANISM],
+    [PUBTATOR_ANNOTATION_TYPE.CHEMICAL, HINT_TYPE.CHEMICAL],
+    [PUBTATOR_ANNOTATION_TYPE.DISEASE, HINT_TYPE.DISEASE],
+    [PUBTATOR_ANNOTATION_TYPE.CELL_LINE, HINT_TYPE.CELL_LINE],
   ]);
   const database2Xref = new Map([
+    [PUBTATOR_ANNOTATION_TYPE.GENE, COLLECTIONS.NCBI_GENE],
     [PUBTATOR_DATABASE.ncbi_taxonomy, COLLECTIONS.NCBI_TAXONOMY],
+    [PUBTATOR_DATABASE.ncbi_mesh, COLLECTIONS.MESH],
+    [PUBTATOR_DATABASE.cvcl, COLLECTIONS.CELLOSAURUS],
   ]);
 
   /**

src/util/registry.js

@@ -1,17 +1,36 @@
 const COLLECTIONS = Object.freeze({
-    PUBMED: {
-        dbname: 'PubMed',
-        dbPrefix: 'pubmed'
-    },
-    /** 
-     * The Taxonomy Database is a curated classification and nomenclature for all of the organisms 
-     * in the public sequence databases. This currently represents about 10% of the described species of life on the planet.
-    */
-    NCBI_TAXONOMY: { 
-        dbname: 'NCBI Taxonomy',
-        dbPrefix: 'NCBITaxon'
-    }
+  PUBMED: {
+    dbname: 'PubMed',
+    dbPrefix: 'pubmed',
+  },
+  NCBI_GENE: {
+    dbName: 'NCBI Gene',
+    dbPrefix: 'NCBIGene',
+  },
+  /**
+   * The Taxonomy Database is a curated classification and nomenclature for all of the organisms
+   * in the public sequence databases. This currently represents about 10% of the described species of life on the planet.
+   */
+  NCBI_TAXONOMY: {
+    dbname: 'NCBI Taxonomy',
+    dbPrefix: 'NCBITaxon',
+  },
+  MESH: {
+    dbName: 'MeSH',
+    dbPrefix: 'mesh',
+  },
+  CHEBI: {
+    dbName: 'ChEBI',
+    dbPrefix: 'CHEBI',
+  },
+  CELLOSAURUS: {
+    dbName: 'CELLOSAURUS',
+    dbPrefix: 'cellosaurus',
+  },
+  UNIPROT: {
+    dbName: 'UniProt Knowledgebase',
+    dbPrefix: 'uniprot',
+  },
 });
 
 export { COLLECTIONS };
-