@sjcrh/proteinpaint-server 2.177.0 → 2.177.1-0

package/dataset/termdb.test.js

@@ -373,6 +373,10 @@ function termdb_test_default() {
       ssGSEA: {
         file: "files/hg38/TermdbTest/rnaseq/TermdbTest.ssgsea.h5"
       },
+      dnaMethylation: {
+        file: "files/hg38/TermdbTest/dnaMeth.h5",
+        unit: "Average Beta Value"
+      },
       topVariablyExpressedGenes: {
         src: "native"
       },

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@sjcrh/proteinpaint-server",
-  "version": "2.177.0",
+  "version": "2.177.1-0",
   "type": "module",
   "description": "a genomics visualization tool for exploring a cohort's genotype and phenotype data",
   "main": "src/app.js",
@@ -62,11 +62,11 @@
   },
   "dependencies": {
     "@sjcrh/augen": "2.143.0",
-    "@sjcrh/proteinpaint-python": "2.174.0",
+    "@sjcrh/proteinpaint-python": "2.177.1-0",
     "@sjcrh/proteinpaint-r": "2.152.1-0",
-    "@sjcrh/proteinpaint-rust": "2.177.0",
-    "@sjcrh/proteinpaint-shared": "2.177.0",
-    "@sjcrh/proteinpaint-types": "2.177.0",
+    "@sjcrh/proteinpaint-rust": "2.177.1-0",
+    "@sjcrh/proteinpaint-shared": "2.177.1-0",
+    "@sjcrh/proteinpaint-types": "2.177.1-0",
     "@types/express": "^5.0.0",
     "@types/express-session": "^1.18.1",
     "better-sqlite3": "^12.4.1",

package/routes/termdb.chat.js

@@ -870,28 +870,72 @@ function removeLastOccurrence(str, word) {
     return str.slice(0, index) + str.slice(index + word.length);
   }
 }
+function sortSameCategoricalFilterKeys(filters, ds) {
+  let html = "";
+  const keys = filters.map((f) => f.term);
+  if (new Set(keys).size == keys.length) return { filters, html };
+  const seen = /* @__PURE__ */ new Set();
+  const categorical_filter_terms_with_multiple_fields = /* @__PURE__ */ new Set();
+  for (const item of filters) {
+    if (seen.has(item.term)) categorical_filter_terms_with_multiple_fields.add(item.term);
+    else seen.add(item.term);
+  }
+  const multiple_fields_keys = [];
+  for (const key of categorical_filter_terms_with_multiple_fields) {
+    const term = ds.cohort.termdb.q.termjsonByOneid(key);
+    if (!term) {
+      html += "invalid filter id:" + key;
+    } else {
+      if (term.type == "categorical") {
+        const multiple_fields = filters.filter((x) => x.term == key);
+        multiple_fields_keys.push({ key, categories: multiple_fields.map((f) => f.category) });
+      }
+    }
+  }
+  const sorted_filter = [];
+  const seen2 = /* @__PURE__ */ new Set();
+  for (const f of filters) {
+    const repeated_term = multiple_fields_keys.find((x) => x.key == f.term);
+    if (!repeated_term) {
+      sorted_filter.push(f);
+    } else {
+      if (!seen2.has(f.term)) {
+        const new_filter_term = {
+          term: f.term,
+          category: repeated_term.categories
+        };
+        seen2.add(f.term);
+        sorted_filter.push(new_filter_term);
+      }
+    }
+  }
+  return { filters: sorted_filter, html };
+}
 function validate_filter(filters, ds, group_name) {
   if (!Array.isArray(filters)) throw "filter is not array";
-  let filter_result = { html: "" };
-  if (filters.length <= 2) {
-    filter_result = generate_filter_term(filters, ds);
+  const sorted_filters = sortSameCategoricalFilterKeys(filters, ds);
+  let filter_result = { html: sorted_filters.html };
+  if (sorted_filters.filters.length <= 2) {
+    const generated = generate_filter_term(sorted_filters.filters, ds);
+    filter_result.simplefilter = generated.simplefilter;
+    filter_result.html += generated.html;
   } else {
-    if (filters.length > num_filter_cutoff) {
-      filter_result.html = "For now, the maximum number of filter terms supported through the chatbot is " + num_filter_cutoff;
+    if (sorted_filters.filters.length > num_filter_cutoff) {
+      filter_result.html += "For now, the maximum number of filter terms supported through the chatbot is " + num_filter_cutoff;
       if (group_name.length > 0) {
-        filter_result.html += " . The number of filter terms for group " + group_name + " is " + filters.length + "\n";
+        filter_result.html += " . The number of filter terms for group " + group_name + " is " + sorted_filters.filters.length + "\n";
       } else {
-        filter_result.html += "The number of filter terms for this query is " + filters.length;
+        filter_result.html += "The number of filter terms for this query is " + sorted_filters.filters.length;
       }
     } else {
-      for (let i = 0; i < filters.length - 1; i++) {
+      for (let i = 0; i < sorted_filters.filters.length - 1; i++) {
         const filter_lst = [];
         if (i == 0) {
-          filter_lst.push(filters[i]);
+          filter_lst.push(sorted_filters.filters[i]);
         } else {
           filter_lst.push(filter_result.simplefilter);
         }
-        filter_lst.push(filters[i + 1]);
+        filter_lst.push(sorted_filters.filters[i + 1]);
         filter_result = generate_filter_term(filter_lst, ds);
       }
     }
@@ -913,19 +957,35 @@ function generate_filter_term(filters, ds) {
           localfilter.join = f.join;
         }
         if (term.type == "categorical") {
-          let cat;
-          for (const ck in term.values) {
-            if (ck == f.category) cat = ck;
-            else if (term.values[ck].label == f.category) cat = ck;
-          }
-          if (!cat) invalid_html += "invalid category from " + JSON.stringify(f);
-          localfilter.lst.push({
-            type: "tvs",
-            tvs: {
-              term,
-              values: [{ key: cat }]
+          if (Array.isArray(f.category)) {
+            const categories = [];
+            for (const category of f.category) {
+              const cat = findCategoryKey(term.values, category);
+              if (!cat) invalid_html += "invalid category from " + JSON.stringify(f);
+              else {
+                categories.push({ key: cat });
+              }
             }
-          });
+            localfilter.lst.push({
+              type: "tvs",
+              tvs: {
+                term,
+                values: categories
+              }
+            });
+          } else {
+            const cat = findCategoryKey(term.values, f.category);
+            if (!cat) invalid_html += "invalid category from " + JSON.stringify(f);
+            else {
+              localfilter.lst.push({
+                type: "tvs",
+                tvs: {
+                  term,
+                  values: [{ key: cat }]
+                }
+              });
+            }
+          }
         } else if (term.type == "float" || term.type == "integer") {
           const numeric = {
             type: "tvs",
@@ -958,6 +1018,12 @@ function generate_filter_term(filters, ds) {
   }
   return { simplefilter: localfilter, html: invalid_html };
 }
+function findCategoryKey(termValues, category) {
+  for (const ck in termValues) {
+    if (ck === category || termValues[ck].label === category) return ck;
+  }
+  return void 0;
+}
 async function parse_geneset_db(genedb) {
   let genes_list = [];
   const db = new Database(genedb);

package/routes/termdb.config.js

@@ -189,6 +189,9 @@ function addNonDictionaryQueries(c, ds, genome) {
   if (q.geneExpression) {
     q2.geneExpression = { unit: q.geneExpression.unit };
   }
+  if (q.dnaMethylation) {
+    q2.dnaMethylation = { unit: q.dnaMethylation.unit };
+  }
   if (q.ld) {
     q2.ld = structuredClone(q.ld);
   }
@@ -264,6 +267,7 @@ function getAllowedTermTypes(ds) {
   if (ds.queries?.geneExpression) typeSet.add(TermTypes.GENE_EXPRESSION);
   if (ds.queries?.metaboliteIntensity) typeSet.add(TermTypes.METABOLITE_INTENSITY);
   if (ds.queries?.ssGSEA) typeSet.add(TermTypes.SSGSEA);
+  if (ds.queries?.dnaMethylation) typeSet.add(TermTypes.DNA_METHYLATION);
   if (ds.cohort.termdb.numericTermCollections) typeSet.add("termCollection");
   return [...typeSet];
 }