@datagrok/proteomics 1.0.0 → 1.2.0

package/package.json

@@ -1,62 +1,87 @@
-{
-  "name": "@datagrok/proteomics",
-  "friendlyName": "Proteomics",
-  "version": "1.0.0",
-  "description": "Proteomics package for Datagrok",
-  "author": {
-    "name": "Davit Rizhinashvili",
-    "email": "drizhinashvili@datagrok.ai"
-  },
-  "dependencies": {
-    "@datagrok-libraries/bio": "^5.60.1",
-    "@datagrok-libraries/utils": "^4.6.5",
-    "@typescript-eslint/eslint-plugin": "^8.39.0",
-    "@typescript-eslint/parser": "^8.39.0",
-    "cash-dom": "^8.1.5",
-    "css-loader": "^7.1.2",
-    "datagrok-api": "^1.26.0",
-    "dayjs": "^1.11.13",
-    "eslint": "^8.57.1",
-    "eslint-config-google": "^0.14.0",
-    "@datagrok-libraries/test": "^1.1.0"
-  },
-  "devDependencies": {
-    "@types/wu": "^2.1.44",
-    "datagrok-tools": "^5.0.0",
-    "ts-loader": "^9.5.4",
-    "typescript": "^5.9.3",
-    "webpack": "^5.95.0",
-    "webpack-cli": "^5.1.4"
-  },
-  "scripts": {
-    "link-all": "grok link",
-    "debug-proteomics": "webpack && grok publish",
-    "release-proteomics": "webpack && grok publish --release",
-    "build-proteomics": "webpack",
-    "build": "grok api && grok check && webpack",
-    "debug-proteomics-dev": "webpack && grok publish dev",
-    "release-proteomics-dev": "webpack && grok publish dev --release",
-    "debug-proteomics-local": "webpack && grok publish local",
-    "release-proteomics-local": "webpack && grok publish local --release",
-    "debug-proteomics-localhost": "webpack && grok publish localhost",
-    "release-proteomics-localhost": "webpack && grok publish localhost --release",
-    "debug-proteomics-release": "webpack && grok publish release",
-    "release-proteomics-release": "webpack && grok publish release --release",
-    "debug-proteomics-test": "webpack && grok publish test",
-    "release-proteomics-test": "webpack && grok publish test --release"
-  },
-  "canEdit": [
-    "Developers"
-  ],
-  "canView": [
-    "All users"
-  ],
-  "repository": {
-    "type": "git",
-    "url": "https://github.com/datagrok-ai/Proteomics.git",
-    "directory": "."
-  },
-  "overrides": {
-    "datagrok-api": "$datagrok-api"
-  }
-}
+{
+  "name": "@datagrok/proteomics",
+  "friendlyName": "Proteomics",
+  "version": "1.2.0",
+  "author": {
+    "name": "Ed Jaeger",
+    "email": "ejaeger@datagrok.ai"
+  },
+  "description": "Mass spectrometry-based proteomics data analysis: import, QC, normalization, imputation, differential expression, and visualization.",
+  "repository": {
+    "type": "git",
+    "url": "https://github.com/datagrok-ai/public.git",
+    "directory": "packages/Proteomics"
+  },
+  "dependencies": {
+    "@datagrok-libraries/bio": "^5.61.3",
+    "@datagrok-libraries/math": "^1.2.6",
+    "@datagrok-libraries/ml": "^6.10.7",
+    "@datagrok-libraries/statistics": "^1.2.12",
+    "@datagrok-libraries/utils": "^4.6.9",
+    "@datagrok-libraries/test": "^1.1.0",
+    "cash-dom": "^8.1.5",
+    "datagrok-api": "^1.25.0",
+    "rxjs": "^6.5.5",
+    "wu": "^2.1.0"
+  },
+  "devDependencies": {
+    "@types/wu": "^2.1.44",
+    "@typescript-eslint/eslint-plugin": "^8.8.1",
+    "@typescript-eslint/parser": "^8.8.1",
+    "css-loader": "^7.1.2",
+    "eslint": "^8.57.1",
+    "eslint-config-google": "^0.14.0",
+    "style-loader": "^4.0.0",
+    "ts-loader": "^9.5.1",
+    "typescript": "^5.6.3",
+    "webpack": "^5.95.0",
+    "webpack-cli": "^5.1.4"
+  },
+  "scripts": {
+    "link-api": "npm link datagrok-api",
+    "link-all": "npm link datagrok-api @datagrok-libraries/utils @datagrok-libraries/math @datagrok-libraries/ml @datagrok-libraries/bio @datagrok-libraries/statistics",
+    "build": "grok api && grok check --soft && webpack",
+    "debug-proteomics": "grok publish",
+    "release-proteomics": "grok publish --release",
+    "debug-proteomics-local": "grok publish local",
+    "release-proteomics-local": "grok publish local --release",
+    "lint": "eslint \"./src/**/*.ts\"",
+    "lint-fix": "eslint \"./src/**/*.ts\" --fix",
+    "test": "grok test",
+    "test-local": "grok test --host localhost",
+    "build-all": "npm --prefix ./../../js-api run build && npm --prefix ./../../libraries/utils run build && npm --prefix ./../../libraries/math run build && npm --prefix ./../../libraries/ml run build && npm --prefix ./../../libraries/bio run build && npm --prefix ./../../libraries/statistics run build && npm run build"
+  },
+  "canEdit": [
+    "Administrators"
+  ],
+  "canView": [
+    "All users"
+  ],
+  "category": "Bioinformatics",
+  "properties": [
+    {
+      "name": "reviewSpaceName",
+      "description": "Umbrella Space that shared review snapshots are published under. Each target gets a child Space inside it.",
+      "propertyType": "string",
+      "defaultValue": "Proteomics-Reviews",
+      "nullable": false
+    },
+    {
+      "name": "reviewNamePrefix",
+      "description": "Prefix for shared review project names. The structural '-<target>-v<version>-<date>' suffix stays fixed so re-share version detection keeps working.",
+      "propertyType": "string",
+      "defaultValue": "Proteomics-Review",
+      "nullable": false
+    },
+    {
+      "name": "verifyPublishedDashboard",
+      "description": "After publishing, re-open the shared dashboard to verify it survives a reload. Safer for client deliverables; turn off for faster demo shares.",
+      "propertyType": "bool",
+      "defaultValue": true,
+      "nullable": false
+    }
+  ],
+  "overrides": {
+    "datagrok-api": "$datagrok-api"
+  }
+}

package/scripts/deqms_de.R

@@ -0,0 +1,60 @@
+#name: deqmsDE
+#description: DEqMS peptide-count-weighted differential expression
+#language: r
+#environment: channels: [conda-forge, bioconda], dependencies: [bioconductor-limma, bioconductor-deqms]
+#input: dataframe exprDf
+#input: int nGroup1
+#input: dataframe peptideDf
+#input: double fcThreshold = 1.0
+#input: double pThreshold = 0.05
+#output: dataframe result
+
+# exprDf columns are named s1, s2, ... with group1 first, then group2
+nTotal <- ncol(exprDf)
+exprMat <- as.matrix(exprDf)
+rownames(exprMat) <- seq_len(nrow(exprMat))
+counts <- as.numeric(peptideDf[[1]])
+
+# Design matrix (same as limma_de.R)
+group <- factor(c(rep("ctrl", nGroup1), rep("treat", nTotal - nGroup1)),
+                levels = c("ctrl", "treat"))
+design <- model.matrix(~ group)
+
+# Limma pipeline (required by DEqMS as base)
+library(limma)
+fit <- lmFit(exprMat, design)
+fit <- eBayes(fit)
+
+hasDeqms <- suppressWarnings(require(DEqMS, quietly = TRUE))
+
+if (hasDeqms) {
+  # DEqMS: peptide-count-weighted variance estimation
+  fit$count <- counts
+  fit <- spectraCounteBayes(fit)
+  # outputResult() returns rows SORTED by significance, NOT input order. The
+  # original 1-based input index is preserved in rownames; carry it through so
+  # the client aligns stats to proteins by key, never by row position.
+  tt <- outputResult(fit, coef = 2)
+
+  result <- data.frame(
+    row = as.integer(rownames(tt)),
+    log2FC = tt$logFC,
+    p.value = tt$sca.P.Value,
+    adj.p.value = tt$sca.adj.pval,
+    significant = (abs(tt$logFC) >= fcThreshold) & (tt$sca.adj.pval <= pThreshold),
+    check.names = FALSE
+  )
+} else {
+  # Fallback: use limma topTable (DEqMS not installed)
+  warning("DEqMS not available, using limma instead")
+  tt <- topTable(fit, coef = 2, number = Inf, sort.by = "none")
+
+  result <- data.frame(
+    row = as.integer(rownames(tt)),
+    log2FC = tt$logFC,
+    p.value = tt$P.Value,
+    adj.p.value = tt$adj.P.Val,
+    significant = (abs(tt$logFC) >= fcThreshold) & (tt$adj.P.Val <= pThreshold),
+    check.names = FALSE
+  )
+}

package/scripts/limma_de.R

@@ -0,0 +1,42 @@
+#name: limmaDE
+#description: Differential expression via limma moderated t-test
+#language: r
+#environment: channels: [conda-forge, bioconda], dependencies: [bioconductor-limma]
+#input: dataframe exprDf
+#input: int nGroup1
+#input: double fcThreshold = 1.0
+#input: double pThreshold = 0.05
+#output: dataframe result
+
+# exprDf columns are named s1, s2, ... with group1 first, then group2
+nTotal <- ncol(exprDf)
+exprMat <- as.matrix(exprDf)
+nRows <- nrow(exprMat)
+# 1-based input row index, carried through to the result so the client aligns
+# stats to proteins by key — never by row position (some R outputs reorder).
+rownames(exprMat) <- seq_len(nRows)
+g1Idx <- 1:nGroup1
+g2Idx <- (nGroup1 + 1):nTotal
+
+hasLimma <- suppressWarnings(require(limma, quietly = TRUE))
+
+if (hasLimma) {
+  # Limma empirical Bayes moderated t-test
+  group <- factor(c(rep("ctrl", nGroup1), rep("treat", nTotal - nGroup1)),
+                  levels = c("ctrl", "treat"))
+  design <- model.matrix(~ group)
+  fit <- lmFit(exprMat, design)
+  fit <- eBayes(fit)
+  tt <- topTable(fit, coef = 2, number = Inf, sort.by = "none")
+
+  result <- data.frame(
+    row = as.integer(rownames(tt)),
+    log2FC = tt$logFC,
+    p.value = tt$P.Value,
+    adj.p.value = tt$adj.P.Val,
+    significant = (abs(tt$logFC) >= fcThreshold) & (tt$adj.P.Val <= pThreshold),
+    check.names = FALSE
+  )
+} else {
+  stop("limma package is not available")
+}

package/scripts/vsn_normalize.R

@@ -0,0 +1,19 @@
+#name: VsnNormalize
+#description: Variance-stabilizing normalization via vsn
+#language: r
+#environment: channels: [conda-forge, bioconda], dependencies: [bioconductor-vsn]
+#input: dataframe exprDf
+#output: dataframe result
+
+# exprDf columns are named s1, s2, ... containing raw (non-log2) intensities
+# VSN performs a generalized log2 transformation that stabilizes variance
+
+library(vsn)
+
+exprMat <- as.matrix(exprDf)
+
+# Run VSN normalization (returns glog2-transformed values)
+vsnFit <- vsn::justvsn(exprMat)
+
+result <- as.data.frame(vsnFit)
+colnames(result) <- colnames(exprDf)