@datagrok/sequence-translator 1.0.1 → 1.0.2

package/package.json

@@ -1,7 +1,7 @@
 {
   "name": "@datagrok/sequence-translator",
   "friendlyName": "Sequence Translator",
-  "version": "1.0.1",
+  "version": "1.0.2",
   "description": "SequenceTranslator is a [package](https://datagrok.ai/help/develop/develop#packages) for the [Datagrok](https://datagrok.ai) platform, used to translate [oligonucleotide](https://en.wikipedia.org/wiki/Oligonucleotide) sequences between [different representations](https://github.com/datagrok-ai/public/tree/master/packages/SequenceTranslator#sequence-representations).",
   "repository": {
     "type": "git",
@@ -34,7 +34,8 @@
     "test-local": "set HOST=localhost && jest"
   },
   "sources": [
-    "css/style.css"
+    "css/style.css",
+    "common/openchemlib-full.js"
   ],
   "devDependencies": {
     "@typescript-eslint/eslint-plugin": "^4.29.1",

package/src/package.ts

@@ -278,7 +278,7 @@ const weightsObj: {[code: string]: number} = {};
 for (const synthesizer of Object.keys(map)) {
   for (const technology of Object.keys(map[synthesizer])) {
     for (const code of Object.keys(map[synthesizer][technology]))
-      weightsObj[code] = map[synthesizer][technology][code].weight;
+      weightsObj[code] ?? map[synthesizer][technology][code].weight;
   }
 }
 for (const [key, value] of Object.entries(MODIFICATIONS))

package/src/structures-works/converters.ts

@@ -1,3 +1,23 @@
+import {lcmsToGcrs} from './map';
+import * as DG from 'datagrok-api/dg';
+
+//name: gcrsToLcms
+//input: string nucleotides {semType: GCRS}
+//output: string result {semType: LCMS}
+export function gcrsToLcms(sequence: string): string {
+  const df = DG.DataFrame.fromCsv(lcmsToGcrs);
+  const arr1 = df.getCol('GCRS').toList();
+  const arr2 = df.getCol('LCMS').toList();
+  const obj: {[i: string]: string} = {};
+  arr1.forEach((element, index) => obj[element] = arr2[index]);
+  for (let i = 0; i < arr1.length; i++) {
+    arr1[i] = arr1[i].replace('(', '\\(');
+    arr1[i] = arr1[i].replace(')', '\\)');
+  }
+  const regExp = new RegExp('(' + arr1.join('|') + ')', 'g');
+  return sequence.replace(regExp, function(code) {return obj[code];});
+}
+
 //name: asoGapmersNucleotidesToBioSpring
 //input: string nucleotides {semType: DNA nucleotides}
 //output: string result {semType: BioSpring / Gapmers}

package/src/structures-works/map.ts

@@ -1,9 +1,13 @@
+import * as DG from 'datagrok-api/dg';
+import {getAllCodesOfSynthesizer} from './sequence-codes-tools';
+
 export const SYNTHESIZERS = {
   RAW_NUCLEOTIDES: 'Raw Nucleotides',
   BIOSPRING: 'BioSpring Codes',
   GCRS: 'Janssen GCRS Codes',
   AXOLABS: 'Axolabs Codes',
   MERMADE_12: 'Mermade 12',
+  LCMS: 'LCMS',
 };
 export const TECHNOLOGIES = {
   DNA: 'DNA',
@@ -52,7 +56,7 @@ export const MODIFICATIONS: {[index: string]: {molecularWeight: number, left: st
 export const stadardPhosphateLinkSmiles = 'OP(=O)(O)O';
 export const map: {[synthesizer: string]:
   {[technology: string]: {[code: string]:
-    {'name': string, 'weight': number, 'normalized': string, 'SMILES': string}}}} = {
+    {'name'?: string, 'weight'?: number, 'normalized'?: string, 'SMILES': string}}}} = {
       'Raw Nucleotides': {
         'DNA': {
           'A': {
@@ -458,6 +462,7 @@ export const map: {[synthesizer: string]:
             'SMILES': 'OC[C@H]1O[C@@H](N2C3N=C(N)NC(=O)C=3N=C2)[C@H](OC)[C@@H]1O',
           },
         },
+        'Others': {},
       },
       'Mermade 12': {
         'For 2\'-OMe and 2\'-F modified siRNA': {
@@ -559,4 +564,164 @@ export const map: {[synthesizer: string]:
           },
         },
       },
+      // 'LCMS': {
+      //   'For 2\'-OMe and 2\'-F modified siRNA': {
     };
+
+export const lcmsToGcrs = `LCMS, GCRS
+A, A
+C, C
+/5mC/, (5m)C
+G, G
+T, T
+rA, rA
+rC, rC
+rG, rG
+rU, rU
+mA, mA
+mC, mC
+/5mmC/, (5m)mC
+mG, mG
+mU, mU
+fA, fA
+fC, fC
+/5mfC/, (5m)fC
+fG, fG
+fU, fU
+/afA/, afA
+/afC/, afC
+/afG/, afG
+/afU/, afU
++A, lna A
++C, lna C
++G, lna G
++T, lna T
+/moeA/, moeA
+/moeC/, moeC
+/5mmoeC/, (5m)moeC
+/moeG/, moeG
+/moeT/, moeT
+/moeU/, moeU
+/xA/, Anp
+/xC/, Cnp
+/x5mC/, (5m)Cnp
+/xG/, Gnp
+/xT/, Tnp
+/xrA/, rAnp
+/xrC/, rCnp
+/xrG/, rGnp
+/xrU/, rUnp
+/xmA/, mAnp
+/xmC/, mCnp
+/x5mmC/, (5m)mCnp
+/xmG/, mGnp
+/xmU/, mUnp
+/xfA/, fAnp
+/xfC/, fCnp
+/xfG/, fGnp
+/xfT/, fTnp
+/xfU/, fUnp
+/xafA/, afAnp
+/xafC/, afCnp
+/xafG/, afGnp
+/xafU/, afUnp
+/xeA/, eAnp
+/xeC/, eCnp
+/xeG/, eGnp
+/xeU/, eUnp
+/xmoeA/, moeAnp
+/xmoeC/, moeCnp
+/x5mmoeC/, (5m)moeCnp
+/xmoeG/, moeGnp
+/xmoeU/, moeUnp
+/UNA-A/, (UNA-A)
+/UNA-C/, (UNA-C)
+/UNA-G/, (UNA-G)
+/UNA-T/, (UNA-T)
+/UNA-U/, (UNA-U)
+/GNA-A/, (GNA-A)
+/GNA-C/, (GNA-C)
+/GNA-G/, (GNA-G)
+/GNA-T/, (GNA-T)
+/GNA-U/, (GNA-U)
+/5CholTEG/, (5-CholTEG)
+/3CholTEG/, (TEGChol-3)
+/Toco/, Toco
+/Palm/, Palm
+/GalNAc/, GalNAc
+/GalNAc2/, GalNAc2
+/GalNAc3/, GalNAc3
+/GalNAc6/, GalNAc6
+/GalNAc7/, GalNAc7
+/GalNAc9/, GalNAc9
+/GalNAc14/, GalNAc14
+/NAG37/, NAG37
+/HEG/, (HEG)
+/TEG/, (TEG)
+/AmmC6/, (NHC6)
+/AmmC7/, (NHC7)
+/AmmC12/, (NHC12)
+/invAb/, (invabasic)
+/invdT/, (invdT)
+/VPmU/, (vinu)
+*, ps
+/2-C16U/, 2-C16U 
+/2-C18w9U/, 2-C18w9U
+/JDi-Palm/, JDi-Palm
+/J2-CONC16U/, J2-CONC16U
+/J2-C3NC16U/, J2-C3NC16U
+/J-C15Ada/, J-C15Ada
+/J-2C15AdaU/, J-2C15AdaU
+/J-C16NC6/, J-C16NC6
+/R2-C6NH-U/, R2-C6NH-U
+/J-M1/, J-M1
+/J-B1/, J-B1
+/J-B2/, J-B2
+/J-M2/, J-M2
+/2-C16C/, 2-C16C
+/2-C16A/, 2-C16A
+/2-C16G/, 2-C16G
+/R2-C6NH-G/, R2-C6NH-G
+/R2-C6NH-C/, R2-C6NH-C
+/J2-CONC16A/, J2-CONC16A
+/J2-CONC16C/, J2-CONC16C
+/J2-CONC16G/, J2-CONC16G
+/J2-C15AdaC/, J2-C15AdaC
+/J2-M2U/, J2-M2U
+/J2-B2U/, J2-B2U
+/J2-C3NC16C/, J2-C3NC16C
+/J2-C3NC16G/, J2-C3NC16G
+/R2-C6NH-A/, R2-C6NH-A
+/J2-C15AdaA/, J2-C15AdaA
+/J2-C3NC16A/, J2-C3NC16A
+/J-C5-SER-1/, J-C5-SER-1
+/J-C16-SER-1/, J-C16-SER-1
+/J-A2/, J-A2
+/J-A1/, J-A1
+/J2-C15AdaG/, J2-C15AdaG
+/J-C16NAsp/, J-C16NAsp
+/J2-C16NC6U/, J2-C16NC6U
+/J-C5-REBO-1/, J-C5-REBO-1
+/J-C16-REBO-1/, J-C16-REBO-1
+/J-C16-IND-1/, J-C16-IND-1
+/J-C5-IND-1/, J-C5-IND-1
+/J-1C15Ada-2Man/, J-1C15Ada-2Man
+/JG-1C15Ada-23DiMan/, JG-1C15Ada-2,3DiMan
+/J-TriManPC/, J-TriManPC
+/J-triManPO/, J-triManPO
+/J-A4/, J-A4
+/J-Ara-1/, J-Ara-1
+/J-Ara-2/, J-Ara-2
+/J-AcCS/, J-AcCS
+/J-CbCS/, J-CbCS
+/J-MtCD/, J-MtCD`;
+
+function differenceOfTwoArrays(a: string[], b: string[]): string[] {
+  return a.filter((x) => !b.includes(x));
+}
+
+const codesWithSmiles = getAllCodesOfSynthesizer(SYNTHESIZERS.GCRS);
+const allGcrsCodes = DG.DataFrame.fromCsv(lcmsToGcrs).getCol('GCRS').toList();
+export const gcrsCodesWithoutSmiles = differenceOfTwoArrays(allGcrsCodes, codesWithSmiles);
+for (const e of gcrsCodesWithoutSmiles)
+  map[SYNTHESIZERS.GCRS]['Others'][e] = {'SMILES': ''};

package/src/structures-works/sequence-codes-tools.ts

@@ -1,6 +1,4 @@
-// import * as grok from 'datagrok-api/grok';
-// import * as ui from 'datagrok-api/ui';
-// import * as DG from 'datagrok-api/dg';
+
 import {map, SYNTHESIZERS, TECHNOLOGIES, MODIFICATIONS} from './map';
 import {asoGapmersNucleotidesToBioSpring, asoGapmersNucleotidesToGcrs,
   asoGapmersBioSpringToNucleotides, asoGapmersBioSpringToGcrs, asoGapmersGcrsToNucleotides,
@@ -8,7 +6,7 @@ import {asoGapmersNucleotidesToBioSpring, asoGapmersNucleotidesToGcrs,
   siRnaNucleotideToAxolabsSenseStrand, siRnaNucleotidesToGcrs, siRnaBioSpringToNucleotides,
   siRnaBioSpringToAxolabs, siRnaBioSpringToGcrs, siRnaAxolabsToNucleotides,
   siRnaAxolabsToBioSpring, siRnaAxolabsToGcrs, siRnaGcrsToNucleotides,
-  siRnaGcrsToBioSpring, siRnaGcrsToAxolabs, gcrsToNucleotides} from './converters';
+  siRnaGcrsToBioSpring, siRnaGcrsToAxolabs, gcrsToNucleotides, gcrsToLcms} from './converters';
 
 const noTranslationTableAvailable = 'No translation table available';
 export const undefinedInputSequence = 'Type of input sequence is undefined';
@@ -206,11 +204,11 @@ export function isValidSequence(sequence: string, format: string | null): {
   };
 }
 
-function getAllCodesOfSynthesizer(synthesizer: string): string[] {
+export function getAllCodesOfSynthesizer(synthesizer: string): string[] {
   let codes: string[] = [];
   for (const technology of Object.keys(map[synthesizer]))
     codes = codes.concat(Object.keys(map[synthesizer][technology]));
-  return codes.concat(Object.keys(MODIFICATIONS));
+  return codes.concat(Object.keys(MODIFICATIONS)).concat(',');
 }
 
 function getListOfPossibleSynthesizersByFirstMatchedCode(sequence: string): string[] {
@@ -273,6 +271,7 @@ export function convertSequence(sequence: string, output: {
       BioSpring: asoGapmersGcrsToBioSpring(sequence),
       Mermade12: gcrsToMermade12(sequence),
       GCRS: sequence,
+      LCMS: gcrsToLcms(sequence),
     };
   }
   if (output.synthesizer!.includes(SYNTHESIZERS.RAW_NUCLEOTIDES) && output.technology!.includes(TECHNOLOGIES.RNA)) {
@@ -310,6 +309,7 @@ export function convertSequence(sequence: string, output: {
       Axolabs: siRnaGcrsToAxolabs(sequence),
       MM12: gcrsToMermade12(sequence),
       GCRS: sequence,
+      LCMS: gcrsToLcms(sequence),
     };
   }
   if (output.synthesizer!.includes(SYNTHESIZERS.GCRS)) {
@@ -318,6 +318,7 @@ export function convertSequence(sequence: string, output: {
       Nucleotides: gcrsToNucleotides(sequence),
       GCRS: sequence,
       Mermade12: gcrsToMermade12(sequence),
+      LCMS: gcrsToLcms(sequence),
     };
   }
   if (output.synthesizer!.includes(SYNTHESIZERS.MERMADE_12)) {

package/{test-SequenceTranslator-089b6516ed77-2280593f.html → test-SequenceTranslator-b5ff4c12f22d-f3bfb562.html}

@@ -1,4 +1,4 @@
-<html><head><meta charset="utf-8"/><title>SequenceTranslator Test Report. Datagrok version datagrok/datagrok:latest SHA=089b6516ed77. Commit 2280593f.</title><style type="text/css">html,
+<html><head><meta charset="utf-8"/><title>SequenceTranslator Test Report. Datagrok version datagrok/datagrok:latest SHA=b5ff4c12f22d. Commit f3bfb562.</title><style type="text/css">html,
 body {
   font-family: Arial, Helvetica, sans-serif;
   font-size: 1rem;
@@ -229,7 +229,7 @@ header {
   font-size: 1rem;
   padding: 0 0.5rem;
 }
-</style></head><body><div id="jesthtml-content"><header><h1 id="title">SequenceTranslator Test Report. Datagrok version datagrok/datagrok:latest SHA=089b6516ed77. Commit 2280593f.</h1></header><div id="metadata-container"><div id="timestamp">Started: 2022-06-09 16:25:38</div><div id="summary"><div id="suite-summary"><div class="summary-total">Suites (1)</div><div class="summary-passed summary-empty">0 passed</div><div class="summary-failed">1 failed</div><div class="summary-pending summary-empty">0 pending</div></div><div id="test-summary"><div class="summary-total">Tests (1)</div><div class="summary-passed summary-empty">0 passed</div><div class="summary-failed">1 failed</div><div class="summary-pending summary-empty">0 pending</div></div></div></div><div id="suite-1" class="suite-container"><div class="suite-info"><div class="suite-path">/home/runner/work/public/public/packages/SequenceTranslator/src/__jest__/remote.test.ts</div><div class="suite-time warn">13.491s</div></div><div class="suite-tests"><div class="test-result failed"><div class="test-info"><div class="test-suitename"> </div><div class="test-title">TEST</div><div class="test-status">failed</div><div class="test-duration">0.623s</div></div><div class="failureMessages"> <pre class="failureMsg">Error: Evaluation failed: Unable to find JS function "test"
+</style></head><body><div id="jesthtml-content"><header><h1 id="title">SequenceTranslator Test Report. Datagrok version datagrok/datagrok:latest SHA=b5ff4c12f22d. Commit f3bfb562.</h1></header><div id="metadata-container"><div id="timestamp">Started: 2022-06-30 13:09:46</div><div id="summary"><div id="suite-summary"><div class="summary-total">Suites (1)</div><div class="summary-passed summary-empty">0 passed</div><div class="summary-failed">1 failed</div><div class="summary-pending summary-empty">0 pending</div></div><div id="test-summary"><div class="summary-total">Tests (1)</div><div class="summary-passed summary-empty">0 passed</div><div class="summary-failed">1 failed</div><div class="summary-pending summary-empty">0 pending</div></div></div></div><div id="suite-1" class="suite-container"><div class="suite-info"><div class="suite-path">/home/runner/work/public/public/packages/SequenceTranslator/src/__jest__/remote.test.ts</div><div class="suite-time warn">14.488s</div></div><div class="suite-tests"><div class="test-result failed"><div class="test-info"><div class="test-suitename"> </div><div class="test-title">TEST</div><div class="test-status">failed</div><div class="test-duration">0.427s</div></div><div class="failureMessages"> <pre class="failureMsg">Error: Evaluation failed: Unable to find JS function "test"
     at ExecutionContext._evaluateInternal (/home/runner/work/public/public/packages/SequenceTranslator/node_modules/puppeteer/src/common/ExecutionContext.ts:273:13)
     at processTicksAndRejections (internal/process/task_queues.js:97:5)
     at ExecutionContext.evaluate (/home/runner/work/public/public/packages/SequenceTranslator/node_modules/puppeteer/src/common/ExecutionContext.ts:140:12)</pre></div></div></div><div class="suite-consolelog"><div class="suite-consolelog-header">Console Log</div><div class="suite-consolelog-item"><pre class="suite-consolelog-item-origin">    at Object.&lt;anonymous&gt; (/home/runner/work/public/public/packages/SequenceTranslator/src/__jest__/test-node.ts:62:11)