@datagrok/peptides 1.11.3 → 1.13.0

package/src/widgets/mutation-cliffs.ts

@@ -33,8 +33,8 @@ export function mutationCliffsWidget(table: DG.DataFrame, model: PeptidesModel):
     const posColCategories = posCol.categories;
     const posColData = posCol.getRawData();
 
-    for (const aar of currentCell[pos]) {
-      const substitutionsMap = substInfo.get(aar)?.get(pos) as Map<number, type.UTypedArray> | undefined;
+    for (const monomer of currentCell[pos]) {
+      const substitutionsMap = substInfo.get(monomer)?.get(pos) as Map<number, type.UTypedArray> | undefined;
       if (typeof substitutionsMap === 'undefined')
         continue;
 

package/src/widgets/peptides.ts

@@ -16,16 +16,16 @@ import {ALIGNMENT, NOTATION, TAGS as bioTAGS} from '@datagrok-libraries/bio/src/
  * @param {DG.DataFrame} df Working table
  * @param {DG.Column} col Aligned sequence column
  * @return {Promise<DG.Widget>} Widget containing peptide analysis */
-export function analyzePeptidesUI(df: DG.DataFrame, col?: DG.Column<string>):
-  { host: HTMLElement, callback: () => Promise<boolean> } {
+export function analyzePeptidesUI(df: DG.DataFrame, col?: DG.Column<string>): {host: HTMLElement, callback: () => Promise<boolean>} {
   const logoHost = ui.div();
-  // logoHost.style.alignContent = 'center';
   let seqColInput: DG.InputBase | null = null;
   if (typeof col === 'undefined') {
     const sequenceColumns = df.columns.toList().filter((dfCol) => dfCol.semType === DG.SEMTYPE.MACROMOLECULE);
     const potentialCol = DG.Utils.firstOrNull(sequenceColumns);
     if (potentialCol === null)
       throw new Error('Peptides Error: table doesn\'t contain sequence columns');
+    else if (potentialCol.stats.missingValueCount !== 0)
+      grok.shell.info('Sequences column contains missing values. They will be ignored during analysis');
 
     seqColInput = ui.columnInput('Sequence', df, potentialCol, () => {
       const seqCol = seqColInput!.value;
@@ -38,8 +38,10 @@ export function analyzePeptidesUI(df: DG.DataFrame, col?: DG.Column<string>):
         viewer.root.style.setProperty('height', '130px');
         return viewer.root;
       }));
-      //TODO: add when new version of datagrok-api is available
+      if (seqCol.stats.missingValueCount !== 0)
+        grok.shell.info('Sequences column contains missing values. They will be ignored during analysis');
     }, {filter: (col: DG.Column) => col.semType === DG.SEMTYPE.MACROMOLECULE});
+    seqColInput.setTooltip('Macromolecule column in FASTA, HELM or separated format');
   } else if (!(col.getTag(bioTAGS.aligned) === ALIGNMENT.SEQ_MSA) &&
     col.getTag(DG.TAGS.UNITS) !== NOTATION.HELM) {
     return {
@@ -65,10 +67,8 @@ export function analyzePeptidesUI(df: DG.DataFrame, col?: DG.Column<string>):
   }
 
   let scaledCol: DG.Column<number>;
-
-  const defaultActivityColumn: DG.Column<number> | null =
-    df.col('activity') || df.col('IC50') || DG.Utils.firstOrNull(df.columns.numerical);
-  ;
+  const defaultActivityColumn: DG.Column<number> | null = df.col('activity') || df.col('IC50') ||
+    DG.Utils.firstOrNull(df.columns.numerical);
   const histogramHost = ui.div([], {id: 'pep-hist-host'});
 
   const activityScalingMethod = ui.choiceInput(
@@ -88,23 +88,27 @@ export function analyzePeptidesUI(df: DG.DataFrame, col?: DG.Column<string>):
       histogramHost.lastChild?.remove();
       histogramHost.appendChild(hist.root);
     }) as DG.InputBase<C.SCALING_METHODS | null>;
-  activityScalingMethod.setTooltip('Function to apply for each value in activity column');
+  activityScalingMethod.setTooltip('Activity column transformation method');
 
   const activityScalingMethodState = (): void => {
-    activityScalingMethod.enabled = (activityColumnChoice.value ?? false) &&
-      DG.Stats.fromColumn(activityColumnChoice.value!).min > 0;
-    activityScalingMethod.fireChanged();
+    activityScalingMethod.enabled = (activityColumnChoice.value ?? false) && activityColumnChoice.value!.stats.min > 0;
+    activityScalingMethod.value = C.SCALING_METHODS.NONE;
+    if (activityColumnChoice.value!.stats.missingValueCount !== 0)
+      grok.shell.info('Activity column contains missing values. They will be ignored during analysis');
   };
   //TODO: add when new version of datagrok-api is available
   const activityColumnChoice = ui.columnInput('Activity', df, defaultActivityColumn, activityScalingMethodState,
     {filter: (col: DG.Column) => col.type === DG.TYPE.INT || col.type === DG.TYPE.FLOAT});
-  const clustersColumnChoice = ui.columnInput('Clusters', df, null);
+  activityColumnChoice.setTooltip('Numerical activity column');
+  const clustersColumnChoice = ui.columnInput('Clusters', df, null, null);
+  clustersColumnChoice.setTooltip('Optional. Clusters column is used to create Logo Summary Table');
   clustersColumnChoice.nullable = true;
   activityColumnChoice.fireChanged();
   activityScalingMethod.fireChanged();
 
-  const targetColumnChoice = ui.columnInput('Target', df, null, null,
-    {filter: (col: DG.Column) => col.type === DG.TYPE.STRING});
+  const targetColumnChoice = ui.columnInput('Target', df, null, null, {filter: (col: DG.Column) => col.type === DG.TYPE.STRING});
+  targetColumnChoice.setTooltip('Optional. Target represents a unique binding construct for every peptide in the data. ' +
+    'Target can be used to split mutation cliff analysis for peptides specific to a certain set of targets');
   targetColumnChoice.nullable = true;
 
   const inputsList = [activityColumnChoice, activityScalingMethod, clustersColumnChoice, targetColumnChoice];
@@ -128,7 +132,7 @@ export function analyzePeptidesUI(df: DG.DataFrame, col?: DG.Column<string>):
   const inputElements: HTMLElement[] = [ui.inputs(inputsList)];
   $(inputElements[0]).find('label').css('width', 'unset');
   if (typeof col !== 'undefined') {
-    const startBtn = ui.button('Launch SAR', startAnalysisCallback);
+    const startBtn = ui.button('Launch SAR', startAnalysisCallback, '');
     startBtn.style.alignSelf = 'center';
     inputElements.push(startBtn);
     bottomHeight = '215px';
@@ -198,10 +202,14 @@ export async function startAnalysis(activityColumn: DG.Column<number>, peptidesC
     newDf.setTag(C.TAGS.UUID, dfUuid);
     newDf.setTag('monomerType', monomerType);
 
+    const bitset = DG.BitSet.create(currentDf.rowCount,
+      (i) => !activityColumn.isNone(i) && !peptidesCol.isNone(i) && currentDf.filter.get(i));
+
     // Cloning dataframe with applied filter. If filter is not applied, cloning is
     // needed anyway to allow filtering on the original dataframe
-    model = PeptidesModel.getInstance(newDf.clone(currentDf.filter));
-    if (clustersColumn) await model.addLogoSummaryTable();
+    model = PeptidesModel.getInstance(newDf.clone(bitset));
+    if (clustersColumn)
+      await model.addLogoSummaryTable();
     await model.addMonomerPosition();
     await model.addMostPotentResidues();
   } else

package/src/widgets/selection.ts

@@ -0,0 +1,31 @@
+import * as ui from 'datagrok-api/ui';
+import * as DG from 'datagrok-api/dg';
+
+import {PeptidesModel} from '../model';
+
+import wu from 'wu';
+
+export function getSelectionWidget(table: DG.DataFrame, model: PeptidesModel): DG.Widget {
+  const compBitset = model.getCompoundBitset();
+  if (compBitset.trueCount === 0)
+    return new DG.Widget(ui.divText('No compounds selected'));
+  const newTable = DG.DataFrame.create(table.rowCount);
+  newTable.filter.copyFrom(compBitset);
+  const sourceGrid = model.analysisView.grid;
+  const numericalCols = wu(table.columns.numerical);
+  for (let gridColIdx = 1; gridColIdx < sourceGrid.columns.length; gridColIdx++) {
+    const gridCol = sourceGrid.columns.byIndex(gridColIdx)!;
+    if (!gridCol.visible)
+      continue;
+    const sourceCol = gridCol.column!;
+    const sourceColRawData = sourceCol.getRawData();
+    const sourceColCategories = sourceCol.categories;
+    const getValue = numericalCols.some((col) => col.name === sourceCol.name) ? (i: number): number => sourceColRawData[i] :
+      (i: number): string => sourceColCategories[sourceColRawData[i]];
+    const col = newTable.columns.addNewVirtual(gridCol.name, (i) => getValue(i), sourceCol.type as DG.TYPE);
+    for (const [tag, value] of sourceCol.tags)
+      col.setTag(tag, value);
+  }
+  const newGrid = newTable.plot.grid();
+  return new DG.Widget(ui.box(newGrid.root, {style: {width: '100%'}}));
+}

package/src/widgets/settings.ts

@@ -62,12 +62,17 @@ export function getSettingsDialog(model: PeptidesModel): SettingsElements {
   // General pane options
   const activityCol = ui.columnInput(GENERAL_INPUTS.ACTIVITY, model.df,
     model.df.getCol(model.settings.activityColumnName!), () => result.activityColumnName = activityCol.value!.name,
-    {filter: (col: DG.Column) => (col.type === DG.TYPE.FLOAT || col.type === DG.TYPE.INT) && col.name !== C.COLUMNS_NAMES.ACTIVITY_SCALED});
+    {filter: (col: DG.Column) => (col.type === DG.TYPE.FLOAT || col.type === DG.TYPE.INT) &&
+      col.name !== C.COLUMNS_NAMES.ACTIVITY_SCALED && col.stats.missingValueCount === 0});
+  activityCol.setTooltip('Numeric activity column');
   const activityScaling =
     ui.choiceInput(GENERAL_INPUTS.ACTIVITY_SCALING, currentScaling, Object.values(C.SCALING_METHODS),
       () => result.scaling = activityScaling.value as C.SCALING_METHODS) as DG.InputBase<C.SCALING_METHODS>;
+  activityScaling.setTooltip('Activity column transformation method');
   const bidirectionalAnalysis = ui.boolInput(GENERAL_INPUTS.BIDIRECTIONAL_ANALYSIS, currentBidirectional,
     () => result.isBidirectional = bidirectionalAnalysis.value) as DG.InputBase<boolean>;
+  bidirectionalAnalysis.setTooltip('Distinguish between positive and negative mean activity difference in ' +
+    'Monomer-Position and Most Potent Residues viewers');
 
   accordion.addPane(SETTINGS_PANES.GENERAL, () => ui.inputs([activityCol, activityScaling, bidirectionalAnalysis]), true);
   inputs[SETTINGS_PANES.GENERAL] = [activityCol, activityScaling, bidirectionalAnalysis];
@@ -88,6 +93,7 @@ export function getSettingsDialog(model: PeptidesModel): SettingsElements {
   const isDendrogramEnabled = wu(model.analysisView.viewers).some((v) => v.type === VIEWER_TYPE.DENDROGRAM);
   const dendrogram = ui.boolInput(VIEWER_TYPE.DENDROGRAM, isDendrogramEnabled ?? false,
     () => result.showDendrogram = dendrogram.value) as DG.InputBase<boolean>;
+  dendrogram.setTooltip('Show dendrogram viewer');
   dendrogram.enabled = getTreeHelperInstance() !== null;
 
   accordion.addPane(SETTINGS_PANES.VIEWERS, () => ui.inputs([dendrogram]), true);
@@ -100,6 +106,7 @@ export function getSettingsDialog(model: PeptidesModel): SettingsElements {
     result.maxMutations = val;
     maxMutations.addPostfix(val.toString());
   }) as DG.InputBase<number>;
+  maxMutations.setTooltip('Maximum number of mutations between reference and mutated sequences');
   maxMutations.addPostfix((settings.maxMutations ?? 1).toString());
   const minActivityDelta = ui.sliderInput(MUTATION_CLIFFS_INPUTS.MIN_ACTIVITY_DELTA, currentMinActivityDelta, 0,
     100, () => {
@@ -108,6 +115,7 @@ export function getSettingsDialog(model: PeptidesModel): SettingsElements {
       $(minActivityDelta.root).find('label.ui-input-description').remove();
       minActivityDelta.addPostfix(val);
     }) as DG.InputBase<number>;
+  minActivityDelta.setTooltip('Minimum activity difference between reference and mutated sequences');
   minActivityDelta.addPostfix((settings.minActivityDelta ?? 0).toString());
   accordion.addPane(SETTINGS_PANES.MUTATION_CLIFFS, () => ui.inputs([maxMutations, minActivityDelta]), true);
   inputs[SETTINGS_PANES.MUTATION_CLIFFS] = [maxMutations, minActivityDelta];
@@ -131,6 +139,7 @@ export function getSettingsDialog(model: PeptidesModel): SettingsElements {
             delete result.columns;
         }
       }) as DG.InputBase<boolean>;
+    isIncludedInput.setTooltip('Include aggregated column value in tooltips, Logo Summary Table and Distribution panel');
 
     const aggregationInput = ui.choiceInput(COLUMNS_INPUTS.AGGREGATION, (currentColumns)[colName] ?? DG.AGG.AVG,
       Object.values(DG.STATS), () => {
@@ -143,6 +152,7 @@ export function getSettingsDialog(model: PeptidesModel): SettingsElements {
             delete result.columns;
         }
       }) as DG.InputBase<DG.AggregationType>;
+    aggregationInput.setTooltip('Aggregation method');
     $(aggregationInput.root).find('label').css('width', 'auto');
     const inputsRow = ui.inputsRow(col.name, [isIncludedInput, aggregationInput]);
     includedColumnsInputs.push(...[isIncludedInput, aggregationInput]);

package/src/widgets/similarity.ts

@@ -1,39 +0,0 @@
-
-import {sequenceChemSimilarity} from '@datagrok-libraries/bio/src/monomer-works/monomer-utils';
-import {ISeqSplitted} from '@datagrok-libraries/bio/src/utils/macromolecule/types';
-import * as DG from 'datagrok-api/dg';
-
-export function calculateIdentity(template: ISeqSplitted, splitSeqDf: DG.DataFrame): DG.Column<number> {
-  const numPositions = splitSeqDf.columns.length;
-  const positionCols: Uint32Array[] = new Array(numPositions);
-  const positionEmptyCategories: number[] = new Array(numPositions);
-  const categoryIndexesTemplate: number[] = new Array(numPositions);
-
-  for (let posIdx = 0; posIdx < numPositions; ++posIdx) {
-    const posCol = splitSeqDf.columns.byIndex(posIdx);
-    positionCols[posIdx] = posCol.getRawData() as Uint32Array;
-    positionEmptyCategories[posIdx] = posCol.categories.indexOf('');
-    categoryIndexesTemplate[posIdx] = posCol.categories.indexOf(template[posIdx] ?? '');
-  }
-
-  const identityScoresCol = DG.Column.float('Identity', splitSeqDf.rowCount);
-  const identityScoresData = identityScoresCol.getRawData();
-  for (let rowIndex = 0; rowIndex < splitSeqDf.rowCount; ++rowIndex) {
-    identityScoresData[rowIndex] = 0;
-    for (let posIdx = 0; posIdx < template.length; ++posIdx) {
-      const categoryIndex = positionCols[posIdx][rowIndex];
-      if (categoryIndex === categoryIndexesTemplate[posIdx])
-        ++identityScoresData[rowIndex];
-    }
-    identityScoresData[rowIndex] /= template.length;
-  }
-
-  return identityScoresCol;
-}
-
-
-export async function calculateSimilarity(template: ISeqSplitted, splitSeqDf: DG.DataFrame): Promise<DG.Column<number>> {
-  const columns = splitSeqDf.columns.toList() as DG.Column<string>[];
-  const scoresCol = await sequenceChemSimilarity(columns, template);
-  return scoresCol;
-}