@datagrok/bio 1.7.9 → 1.7.13

package/src/utils/convert.ts

@@ -4,7 +4,8 @@ import * as grok from 'datagrok-api/grok';
 import $ from 'cash-dom';
 
 import {Subscription} from 'rxjs';
-import {NotationConverter, NOTATION} from '@datagrok-libraries/bio/src/utils/notation-converter';
+import {NotationConverter} from '@datagrok-libraries/bio/src/utils/notation-converter';
+import {NOTATION} from '@datagrok-libraries/bio/src/utils/units-handler';
 
 
 let convertDialog: DG.Dialog | null = null;
@@ -17,7 +18,7 @@ let convertDialogSubs: Subscription[] = [];
  */
 export function convert(col: DG.Column): void {
   const converter = new NotationConverter(col);
-  const current: NOTATION = converter.sourceNotation;
+  const currentNotation: NOTATION = converter.notation;
   //TODO: read all notations
   const notations = [
     NOTATION.FASTA,
@@ -25,28 +26,30 @@ export function convert(col: DG.Column): void {
     NOTATION.HELM
   ];
   const separatorArray = ['-', '.', '/'];
-  const filteredNotations = notations.filter((e) => e !== current);
+  const filteredNotations = notations.filter((e) => e !== currentNotation);
   const targetNotationInput = ui.choiceInput('Convert to', filteredNotations[0], filteredNotations);
 
   const separatorInput = ui.choiceInput('Separator', separatorArray[0], separatorArray);
 
   // hide the separator input for non-SEPARATOR target notations
-  if (targetNotationInput.value !== NOTATION.SEPARATOR)
-    $(separatorInput.root).hide();
-  else
-    $(separatorInput.root).show();
-
-  targetNotationInput.onChanged(async () => {
+  const toggleSeparator = () => {
     if (targetNotationInput.value !== NOTATION.SEPARATOR)
       $(separatorInput.root).hide();
     else
       $(separatorInput.root).show();
+  };
+
+  // set correct visibility on init
+  toggleSeparator();
+
+  targetNotationInput.onChanged( () => {
+    toggleSeparator();
   });
 
   if (convertDialog == null) {
     convertDialog = ui.dialog('Convert sequence notation')
       .add(ui.div([
-        ui.h1('Current notation: ' + current),
+        ui.h1('Current notation: ' + currentNotation),
         targetNotationInput.root,
         separatorInput.root
       ]))

package/src/utils/utils.ts

@@ -2,7 +2,7 @@ import * as DG from 'datagrok-api/dg';
 import {WebLogo, SplitterFunc} from '@datagrok-libraries/bio/src/viewers/web-logo';
 import * as grok from 'datagrok-api/grok';
 import {
-  CAP_GROUP_NAME, CAP_GROUP_SMILES, jsonSdfMonomerLibDict, MONOMER_SYMBOL,
+  CAP_GROUP_NAME, CAP_GROUP_SMILES, jsonSdfMonomerLibDict, MONOMER_ENCODE_MAX, MONOMER_ENCODE_MIN, MONOMER_SYMBOL,
   RGROUP_ALTER_ID, RGROUP_FIELD, RGROUP_LABEL, SDF_MONOMER_NAME
 } from '../const';
 
@@ -11,6 +11,33 @@ export const HELM_CORE_LIB_MONOMER_SYMBOL = 'symbol';
 export const HELM_CORE_LIB_MOLFILE = 'molfile';
 export const HELM_CORE_FIELDS = ['symbol', 'molfile', 'rgroups', 'name'];
 
+
+export function encodeMonomers(col: DG.Column): DG.Column | null {
+  let encodeSymbol = MONOMER_ENCODE_MIN;
+  const monomerSymbolDict:  { [key: string]: number }= {};
+  const units = col.tags[DG.TAGS.UNITS];
+  const sep = col.getTag('separator');
+  const splitterFunc: SplitterFunc = WebLogo.getSplitter(units, sep);
+  const encodedStringArray = [];
+  for (let i = 0; i < col.length; ++i) {
+    let encodedMonomerStr = '';
+    const monomers = splitterFunc(col.get(i));
+    monomers.forEach(m => {
+      if(!monomerSymbolDict[m]) {
+        if(encodeSymbol > MONOMER_ENCODE_MAX) {
+          grok.shell.error(`Not enougth symbols to encode monomers`);
+          return null;
+        }
+        monomerSymbolDict[m] = encodeSymbol;
+        encodeSymbol++;
+      }
+      encodedMonomerStr += String.fromCodePoint(monomerSymbolDict[m]);
+    })
+    encodedStringArray.push(encodedMonomerStr);
+  }
+  return DG.Column.fromStrings('encodedMolecules', encodedStringArray);
+}
+
 export function getMolfilesFromSeq(col: DG.Column, monomersLibObject: any[]): any[][] | null {
   const units = col.tags[DG.TAGS.UNITS];
   const sep = col.getTag('separator');
@@ -18,7 +45,8 @@ export function getMolfilesFromSeq(col: DG.Column, monomersLibObject: any[]): an
   const monomersDict = createMomomersMolDict(monomersLibObject);
   const molFiles = [];
   for (let i = 0; i < col.length; ++i) {
-    const monomers = splitterFunc(col.get(i));
+    const macroMolecule = col.get(i);
+    const monomers = splitterFunc(macroMolecule);
     const molFilesForSeq = [];
     for (let j = 0; j < monomers.length; ++j) {
       if (monomers[j]) {
@@ -34,6 +62,28 @@ export function getMolfilesFromSeq(col: DG.Column, monomersLibObject: any[]): an
   return molFiles;
 }
 
+export function getMolfilesFromSingleSeq(cell: DG.Cell, monomersLibObject: any[]): any[][] | null {
+  const units = cell.column.tags[DG.TAGS.UNITS];
+  const sep = cell.column!.getTag('separator');
+  const splitterFunc: SplitterFunc = WebLogo.getSplitter(units, sep);
+  const monomersDict = createMomomersMolDict(monomersLibObject);
+  const molFiles = [];
+  const macroMolecule = cell.value;
+  const monomers = splitterFunc(macroMolecule);
+  const molFilesForSeq = [];
+  for (let j = 0; j < monomers.length; ++j) {
+    if (monomers[j]) {
+      if (!monomersDict[monomers[j]]) {
+        grok.shell.warning(`Monomer ${monomers[j]} is missing in HELM library. Structure cannot be created`);
+        return null;
+      }
+      molFilesForSeq.push(JSON.parse(JSON.stringify(monomersDict[monomers[j]])));
+    }
+  }
+  molFiles.push(molFilesForSeq);
+  return molFiles;
+}
+
 export function createMomomersMolDict(lib: any[]): { [key: string]: string | any } {
   const dict: { [key: string]: string | any } = {};
   lib.forEach((it) => {
@@ -79,4 +129,4 @@ export function createJsonMonomerLibFromSdf(table: DG.DataFrame): any {
     resultLib.push(monomer);
   }
   return resultLib;
-}
+}

package/src/widgets/representations.ts

@@ -0,0 +1,54 @@
+import * as grok from 'datagrok-api/grok';
+import * as ui from 'datagrok-api/ui';
+import * as DG from 'datagrok-api/dg';
+import {getMolfilesFromSingleSeq, HELM_CORE_LIB_FILENAME} from '../utils/utils';
+import {getMacroMol} from '../utils/atomic-works';
+
+/**
+ * 3D representation widget of macromolecule.
+ *
+ * @export
+ * @param {DG.Cell} macroMolecule macromolecule cell.
+ * @return {Promise<DG.Widget>} Widget.
+ */
+export async function representationsWidget(macroMolecule: DG.Cell, monomersLibObject: any[]): Promise<DG.Widget> {
+  const pi = DG.TaskBarProgressIndicator.create('Creating 3D view');
+
+  let widgetHost;
+  let molBlock3D = '';
+  try {
+    try {
+      const atomicCodes = getMolfilesFromSingleSeq(macroMolecule, monomersLibObject);
+      const result = await getMacroMol(atomicCodes!);
+      const molBlock2D = result[0];
+      molBlock3D = (await grok.functions.call('Bio:Embed', {molBlock2D})) as string;
+    } catch (e) {
+      console.warn(e);
+    }
+
+    try {
+      molBlock3D = molBlock3D.replaceAll('\\n', '\n');
+      const stringBlob = new Blob([molBlock3D], {type: 'text/plain'});
+      const nglHost = ui.div([], {classes: 'd4-ngl-viewer', id: 'ngl-3d-host'});
+
+      //@ts-ignore
+      const stage = new NGL.Stage(nglHost, {backgroundColor: 'white'});
+      //@ts-ignore
+      stage.loadFile(stringBlob, {ext: 'sdf'}).then(function(comp: NGL.StructureComponent) {
+        stage.setSize(300, 300);
+        comp.addRepresentation('ball+stick');
+        comp.autoView();
+      });
+      const sketch = grok.chem.svgMol(molBlock3D);
+      const panel = ui.divH([sketch]);
+
+      widgetHost = ui.div([panel, nglHost]);
+    } catch (e) {
+      widgetHost = ui.divText('Couldn\'t get peptide structure');
+    }
+  } catch (e) {
+    widgetHost = ui.divText('Couldn\'t get peptide structure');
+  }
+  pi.close();
+  return new DG.Widget(widgetHost);
+}

package/{test-Bio-34f75e5127b8-0a8a5821.html → test-Bio-34f75e5127b8-781e9df9.html}

@@ -1,4 +1,4 @@
-<html><head><meta charset="utf-8"/><title>Bio Test Report. Datagrok version datagrok/datagrok:latest SHA=34f75e5127b8. Commit 0a8a5821.</title><style type="text/css">html,
+<html><head><meta charset="utf-8"/><title>Bio Test Report. Datagrok version datagrok/datagrok:latest SHA=34f75e5127b8. Commit 781e9df9.</title><style type="text/css">html,
 body {
   font-family: Arial, Helvetica, sans-serif;
   font-size: 1rem;
@@ -229,10 +229,8 @@ header {
   font-size: 1rem;
   padding: 0 0.5rem;
 }
-</style></head><body><div id="jesthtml-content"><header><h1 id="title">Bio Test Report. Datagrok version datagrok/datagrok:latest SHA=34f75e5127b8. Commit 0a8a5821.</h1></header><div id="metadata-container"><div id="timestamp">Started: 2022-07-19 13:18:04</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/Bio/src/__jest__/remote.test.ts</div><div class="suite-time warn">189.946s</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">177.683s</div></div><div class="failureMessages"> <pre class="failureMsg">Error: Test result : Bio.MSA.is_correct : TypeError: Cannot read properties of undefined (reading 'split')
-Test result : Bio.splitters.testHelm1 : Error: Expected R(U)P at position 0, got U
-Test result : Bio.splitters.testHelm2 : Error: Expected R(U)P at position 1, got U
-Test result : Bio.splitters.testHelm3 : Error: Expected R(U) at position 1, got U)
+</style></head><body><div id="jesthtml-content"><header><h1 id="title">Bio Test Report. Datagrok version datagrok/datagrok:latest SHA=34f75e5127b8. Commit 781e9df9.</h1></header><div id="metadata-container"><div id="timestamp">Started: 2022-07-27 13:06:16</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/Bio/src/__jest__/remote.test.ts</div><div class="suite-time warn">213.765s</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">199.116s</div></div><div class="failureMessages"> <pre class="failureMsg">Error: Test result : Bio.MSA.is_correct : TypeError: Cannot read properties of undefined (reading 'split')
+Test result : Bio.activityCliffs.activityCliffsOpen : Error: Expected "105 cliffs", got "2362 cliffs"
 
     at /home/runner/work/public/public/packages/Bio/src/__jest__/remote.test.ts:67:20
     at Generator.next (&lt;anonymous&gt;)
@@ -306,12 +304,19 @@ Test result : Bio.detectors.samplesTestSmiles2NegativeSmiles : OK
 Test result : Bio.detectors.samplesTestActivityCliffsNegativeSmiles : OK
 Test result : Bio.detectors.samplesFastaPtPosSequence : OK
 Test result : Bio.detectors.samplesTestCerealNegativeCerealName : OK
+Test result : Bio.detectors.samplesTestSpgi100NegativeStereoCategory : OK
+Test result : Bio.detectors.samplesTestSpgi100NegativeScaffoldNames : OK
+Test result : Bio.detectors.samplesTestSpgi100NegativePrimaryScaffoldName : OK
+Test result : Bio.detectors.samplesTestUnichemSourcesNegativeSrcUrl : OK
+Test result : Bio.detectors.samplesTestUnichemSourcesNegativeBaseIdUrl : OK
 Test result : Bio.MSA.test_table.is_not_empty : OK
 Test result : Bio.sequenceSpace.sequenceSpaceOpens : OK
-Test result : Bio.activityCliffs.activityCliffsOpen : OK
 Test result : Bio.splitters.helm1 : OK
 Test result : Bio.splitters.helm2 : OK
 Test result : Bio.splitters.helm3-multichar : OK
+Test result : Bio.splitters.testHelm1 : OK
+Test result : Bio.splitters.testHelm2 : OK
+Test result : Bio.splitters.testHelm3 : OK
 Test result : Bio.renderers.afterMsa : OK
 Test result : Bio.renderers.afterConvert : OK
 Test result : Bio.converters.testFastaPtToSeparator : OK
@@ -330,10 +335,13 @@ Test result : Bio.converters.testSeparatorPtToHelm : OK
 Test result : Bio.converters.testSeparatorDnaToHelm : OK
 Test result : Bio.converters.testSeparatorRnaToHelm : OK
 Test result : Bio.converters.testSeparatorGapsToHelm : OK
-Test result : Bio.converters.HelmDnaToFasta : OK
-Test result : Bio.converters.HelmRnaToFasta : OK
-Test result : Bio.converters.HelmPtToFasta : OK
-Test result : Bio.converters.HelmDnaToSeparator : OK
-Test result : Bio.converters.HelmRnaToSeparator : OK
-Test result : Bio.converters.HelmPtToSeparator : OK
+Test result : Bio.converters.testHelmDnaToFasta : OK
+Test result : Bio.converters.testHelmRnaToFasta : OK
+Test result : Bio.converters.testHelmPtToFasta : OK
+Test result : Bio.converters.testHelmDnaToSeparator : OK
+Test result : Bio.converters.testHelmRnaToSeparator : OK
+Test result : Bio.converters.testHelmPtToSeparator : OK
+Test result : Bio.converters.testHelmLoneRibose : OK
+Test result : Bio.converters.testHelmLoneDeoxyribose : OK
+Test result : Bio.converters.testHelmLonePhosphorus : OK
 </pre></div></div></div></div></body></html>