@datagrok/bio 1.4.0 → 1.5.1

package/src/utils/multiple-sequence-alignment.ts

@@ -26,37 +26,6 @@ function _fastaToStrings(fasta: string): string[] {
   return fasta.replace(/>sample\d+(\r\n|\r|\n)/g, '').split('\n');
 }
 
-/**
- * Converts aligned sequence to semantic type format.
- *
- * @param {string} seq Source sequence.
- * @return {string} Formatted sequence.
- */
-function _castAligned(seq: string): string {
-  let delimited = '';
-
-  for (const char of seq)
-    delimited += char == '-' ? char : `-${char}`;
-
-  return delimited;
-}
-
-/**
- * Formats a batch of sequences to correspond the semantic type.
- *
- * @param {string[]} alignment List of aligned sequences.
- * @return {string[]} Formatted sequences.
- */
-function _stringsToAligned(alignment: string[]): string[] {
-  const nItems = alignment.length;
-  const aligned = new Array<string>(nItems);
-
-  for (let i = 0; i < nItems; ++i)
-    aligned[i] = _castAligned(alignment[i]);
-
-  return aligned;
-}
-
 /**
  * Runs Aioli environment with kalign tool.
  *
@@ -86,8 +55,8 @@ export async function runKalign(col: DG.Column, isAligned = false) : Promise<DG.
   console.warn(output);
 
   const aligned = _fastaToStrings(buf).slice(0, sequences.length);
-  const alignedCol = DG.Column.fromStrings(`msa(${col.name})`, _stringsToAligned(aligned));
-  alignedCol.setTag(DG.TAGS.UNITS, ''); 
+  const alignedCol = DG.Column.fromStrings(`msa(${col.name})`, aligned);
+  alignedCol.setTag(DG.TAGS.UNITS, '');
   alignedCol.semType = C.SEM_TYPES.Macro_Molecule;
   return alignedCol;
 }

package/src/utils/sequence-activity-cliffs.ts

@@ -0,0 +1,30 @@
+import {IDrawTooltipParams} from '@datagrok-libraries/ml/src/viewers/activity-cliffs';
+import * as DG from 'datagrok-api/dg';
+import * as ui from 'datagrok-api/ui';
+
+export async function sequenceGetSimilarities(col: DG.Column, seq: string): Promise<DG.Column | null> {
+  return null;
+}
+
+export function drawTooltip(params: IDrawTooltipParams) {
+  params.tooltips[params.line.id] = ui.divH([]);
+  const columnNames = ui.divV([
+    ui.divText('sequense'),
+    ui.divText(params.activity.name),
+  ]);
+  columnNames.style.fontWeight = 'bold';
+  columnNames.style.display = 'flex';
+  columnNames.style.justifyContent = 'space-between';
+  params.tooltips[params.line.id].append(columnNames);
+  params.line.mols.forEach((mol: number) => {
+    const seq = ui.divText(params.df.get(params.seqCol.name, mol));
+    const activity = ui.divText(params.df.get(params.activity.name, mol).toFixed(2));
+    activity.style.display = 'flex';
+    activity.style.justifyContent = 'left';
+    activity.style.paddingLeft = '30px';
+    params.tooltips[params.line.id].append(ui.divV([
+      seq,
+      activity,
+    ], {style: {paddingLeft: '5px'}}));
+  });
+}

package/src/utils/sequence-space.ts

@@ -0,0 +1,43 @@
+import * as DG from 'datagrok-api/dg';
+import {AvailableMetrics} from '@datagrok-libraries/ml/src/typed-metrics';
+import {reduceDimensinalityWithNormalization} from '@datagrok-libraries/ml/src/sequence-space';
+import {BitArrayMetrics, StringMetrics} from '@datagrok-libraries/ml/src/typed-metrics';
+import {Matrix} from '@datagrok-libraries/utils/src/type-declarations';
+import BitArray from '@datagrok-libraries/utils/src/bit-array';
+import {ISequenceSpaceParams} from '@datagrok-libraries/ml/src/viewers/activity-cliffs';
+
+export interface ISequenceSpaceResult {
+  distance: Matrix;
+  coordinates: DG.ColumnList;
+}
+
+export async function sequenceSpace(spaceParams: ISequenceSpaceParams): Promise<ISequenceSpaceResult> {
+  let preparedData: any;
+  if (!(spaceParams.seqCol!.tags[DG.TAGS.UNITS] === 'HELM')) {
+    const sep = spaceParams.seqCol.getTag('separator');
+    const sepFinal = sep ? sep === '.' ? '\\\.' : sep : '-';
+    const regex = new RegExp(sepFinal, 'g');
+    if (Object.keys(AvailableMetrics['String']).includes(spaceParams.similarityMetric))
+      preparedData = spaceParams.seqCol.toList().map((v) => v.replace(regex, '')) as string[];
+    else
+      preparedData = spaceParams.seqCol.toList().map((v) => v.replace(regex, '')) as string[];
+  } else {
+    preparedData = spaceParams.seqCol.toList();
+  }
+
+  const sequenceSpaceResult = await reduceDimensinalityWithNormalization(
+    preparedData,
+    spaceParams.methodName,
+    spaceParams.similarityMetric as StringMetrics | BitArrayMetrics,
+    spaceParams.options);
+  const cols: DG.Column[] = spaceParams.embedAxesNames.map(
+    (name, index) => DG.Column.fromFloat32Array(name, sequenceSpaceResult.embedding[index]));
+  return {distance: sequenceSpaceResult.distance, coordinates: new DG.ColumnList(cols)};
+}
+
+
+export function getEmbeddingColsNames(df: DG.DataFrame) {
+  const axes = ['Embed_X', 'Embed_Y'];
+  const colNameInd = df.columns.names().filter((it) => it.includes(axes[0])).length + 1;
+  return axes.map((it) => `${it}_${colNameInd}`);
+}