@sjcrh/proteinpaint-server 2.107.0 → 2.108.1-0

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@sjcrh/proteinpaint-server",
-  "version": "2.107.0",
+  "version": "2.108.1-0",
   "type": "module",
   "description": "a genomics visualization tool for exploring a cohort's genotype and phenotype data",
   "main": "src/app.js",
@@ -60,8 +60,8 @@
   "dependencies": {
     "@sjcrh/augen": "2.87.0",
     "@sjcrh/proteinpaint-rust": "2.99.0",
-    "@sjcrh/proteinpaint-shared": "2.106.0",
-    "@sjcrh/proteinpaint-types": "2.107.0",
+    "@sjcrh/proteinpaint-shared": "2.108.0",
+    "@sjcrh/proteinpaint-types": "2.108.0",
     "@types/express": "^5.0.0",
     "@types/express-session": "^1.18.1",
     "better-sqlite3": "^9.4.1",

package/routes/correlationVolcano.js

@@ -3,7 +3,10 @@ import { getData } from "../src/termdb.matrix.js";
 import run_R from "../src/run_R.js";
 import serverconfig from "../src/serverconfig.js";
 import { mayLog } from "#src/helpers.ts";
+import { stdDev } from "#shared/violin.bins.js";
 import path from "path";
+const minArrayLength = 3;
+const minSD = 0.05;
 const api = {
   endpoint: "termdb/correlationVolcano",
   methods: {
@@ -65,16 +68,31 @@ async function compute(q, ds, genome) {
       vtid2array.get(tw.$id).v2.push(variableValue);
     }
   }
+  const [acceptedVariables, skippedVariables] = Array.from(vtid2array.values()).reduce(
+    ([accepted, skipped], t) => {
+      const grterThanOne = t.v1.length > minArrayLength && t.v2.length > minArrayLength;
+      const significantSD = stdDev(t.v1) > minSD && stdDev(t.v2) > minSD;
+      const v = grterThanOne && significantSD ? accepted : skipped;
+      if (v === accepted)
+        accepted.push(t);
+      if (v === skipped)
+        skipped.push({ tw$id: t.id });
+      return [accepted, skipped];
+    },
+    [[], []]
+  );
+  const result = { skippedVariables, variableItems: [] };
+  if (!acceptedVariables.length)
+    return result;
   const input = {
     method: q.correlationMethod || "pearson",
-    terms: [...vtid2array.values()]
+    terms: acceptedVariables
   };
   const time1 = Date.now();
   const output = {
     terms: JSON.parse(await run_R(path.join(serverconfig.binpath, "utils", "corr.R"), JSON.stringify(input)))
   };
   mayLog("Time taken to run correlation analysis:", Date.now() - time1);
-  const result = { variableItems: [] };
   for (const t of output.terms) {
     const t2 = {
       tw$id: t.id,

package/routes/gdc.maf.js

@@ -101,8 +101,10 @@ async function listMafFiles(q, ds) {
       if (normalTypeName)
         file.sample_types.push(normalTypeName);
     }
+    file.sample_types = [...new Set(file.sample_types)];
     files.push(file);
   }
+  files.sort((a, b) => b.file_size - a.file_size);
   const result = {
     files,
     filesTotal: re.data.pagination.total,

package/routes/termdb.DE.js

@@ -58,12 +58,14 @@ async function run_DE(param, ds, term_results) {
     throw "samplelst.groups[0].values.length<1";
   if (param.samplelst.groups[1].values?.length < 1)
     throw "samplelst.groups[1].values.length<1";
-  param.storage_type = ds.queries.rnaseqGeneCount.storage_type;
   const q = ds.queries.rnaseqGeneCount;
   if (!q)
     return;
   if (!q.file)
     throw "unknown data type for rnaseqGeneCount";
+  if (!q.storage_type)
+    throw "storage_type is not defined";
+  param.storage_type = q.storage_type;
   const group1names = [];
   const conf1_group1 = [];
   for (const s of param.samplelst.groups[0].values) {
@@ -73,9 +75,17 @@ async function run_DE(param, ds, term_results) {
     if (!n)
       continue;
     if (q.allSampleSet.has(n)) {
-      group1names.push(n);
       if (param.tw) {
-        conf1_group1.push(term_results.samples[s.sampleId][param.tw.$id]["value"]);
+        if (term_results.samples[s.sampleId]) {
+          if (param.tw.q.mode == "continuous") {
+            conf1_group1.push(term_results.samples[s.sampleId][param.tw.$id]["value"]);
+          } else {
+            conf1_group1.push(term_results.samples[s.sampleId][param.tw.$id]["key"]);
+          }
+          group1names.push(n);
+        }
+      } else {
+        group1names.push(n);
       }
     } else {
     }
@@ -89,9 +99,17 @@ async function run_DE(param, ds, term_results) {
     if (!n)
       continue;
     if (q.allSampleSet.has(n)) {
-      group2names.push(n);
       if (param.tw) {
-        conf1_group2.push(term_results.samples[s.sampleId][param.tw.$id]["value"]);
+        if (term_results.samples[s.sampleId]) {
+          if (param.tw.q.mode == "continuous") {
+            conf1_group2.push(term_results.samples[s.sampleId][param.tw.$id]["value"]);
+          } else {
+            conf1_group2.push(term_results.samples[s.sampleId][param.tw.$id]["key"]);
+          }
+          group2names.push(n);
+        }
+      } else {
+        group2names.push(n);
       }
     } else {
     }
@@ -115,7 +133,10 @@ async function run_DE(param, ds, term_results) {
   };
   if (param.tw) {
     expression_input.conf1 = [...conf1_group2, ...conf1_group1];
-    expression_input.conf1_type = param.tw.term.type;
+    expression_input.conf1_mode = param.tw.q.mode;
+    if (new Set(expression_input.conf1).size === 1) {
+      throw "Confounding variable has only one value";
+    }
   }
   const sample_size_limit = 8;
   let result;

package/routes/termdb.cluster.js

@@ -9,6 +9,7 @@ import { clusterMethodLst, distanceMethodLst } from "#shared/clustering.js";
 import { getResult as getResultGene } from "#src/gene.js";
 import { TermTypes, NUMERIC_DICTIONARY_TERM } from "#shared/terms.js";
 import { getData } from "#src/termdb.matrix.js";
+import { termType2label } from "#shared/terms.js";
 const api = {
   endpoint: "termdb/cluster",
   methods: {
@@ -73,6 +74,12 @@ async function getResult(q, ds, genome) {
     ;
     ({ term2sample2value, byTermId, bySampleId } = await ds.queries[q.dataType].get(_q));
   }
+  for (const [term, obj] of term2sample2value) {
+    if (Object.keys(obj).length === 0) {
+      term2sample2value.delete(term);
+      delete byTermId[term];
+    }
+  }
   if (term2sample2value.size == 0)
     throw "no data";
   if (term2sample2value.size == 1) {
@@ -106,14 +113,24 @@ async function getNumericDictTermAnnotation(q, ds, genome) {
 }
 async function doClustering(data, q, numCases = 1e3) {
   const sampleSet = /* @__PURE__ */ new Set();
+  let firstTerm = true;
   for (const o of data.values()) {
-    for (const s in o)
-      sampleSet.add(s);
-    if (sampleSet.size >= numCases)
-      break;
+    const currentSampleIds = new Set(Object.keys(o));
+    if (firstTerm) {
+      currentSampleIds.forEach((id) => sampleSet.add(id));
+      firstTerm = false;
+    } else {
+      for (const id of sampleSet) {
+        if (!currentSampleIds.has(id)) {
+          sampleSet.delete(id);
+        }
+      }
+    }
   }
   if (sampleSet.size == 0)
-    throw "termdb.cluster: no samples";
+    throw `termdb.cluster: There are no overlapping tested samples shared across the selected ${termType2label(
+      q.dataType
+    )}`;
   if (!clusterMethodLst.find((i) => i.value == q.clusterMethod))
     throw "Invalid cluster method";
   if (!distanceMethodLst.find((i) => i.value == q.distanceMethod))
@@ -122,7 +139,7 @@ async function doClustering(data, q, numCases = 1e3) {
     matrix: [],
     row_names: [],
     // genes
-    col_names: [...sampleSet],
+    col_names: [...sampleSet].slice(0, numCases),
     // samples
     cluster_method: q.clusterMethod,
     distance_method: q.distanceMethod,
@@ -133,7 +150,7 @@ async function doClustering(data, q, numCases = 1e3) {
     inputData.row_names.push(gene);
     const row = [];
     for (const s of inputData.col_names) {
-      row.push(o[s] || 0);
+      row.push(o[s]);
     }
     inputData.matrix.push(q.zScoreTransformation ? getZscore(row) : row);
   }