@datagrok/bio 2.21.10 → 2.21.12

package/package.json

@@ -5,7 +5,7 @@
     "name": "Leonid Stolbov",
     "email": "lstolbov@datagrok.ai"
   },
-  "version": "2.21.10",
+  "version": "2.21.12",
   "description": "Bioinformatics support (import/export of sequences, conversion, visualization, analysis). [See more](https://github.com/datagrok-ai/public/blob/master/packages/Bio/README.md) for details.",
   "repository": {
     "type": "git",
@@ -44,10 +44,10 @@
   ],
   "dependencies": {
     "@biowasm/aioli": "^3.1.0",
-    "@datagrok-libraries/bio": "^5.53.0",
+    "@datagrok-libraries/bio": "^5.53.3",
     "@datagrok-libraries/chem-meta": "^1.2.7",
     "@datagrok-libraries/math": "^1.2.4",
-    "@datagrok-libraries/ml": "^6.10.0",
+    "@datagrok-libraries/ml": "^6.10.2",
     "@datagrok-libraries/tutorials": "^1.6.1",
     "@datagrok-libraries/utils": "^4.5.7",
     "datagrok-api": "^1.25.0",

package/src/package.ts

@@ -82,7 +82,7 @@ export const _package = new BioPackage(/*{debug: true}/**/);
 // let monomerLib: MonomerLib | null = null;
 
 //name: getMonomerLibHelper
-//description:
+//description: Returns an instance of the monomer library helper
 //output: object result
 export async function getMonomerLibHelper(): Promise<IMonomerLibHelper> {
   return await MonomerLibManager.getInstance();
@@ -377,7 +377,6 @@ export function macromoleculeDifferenceCellRenderer(): MacromoleculeDifferenceCe
 
 //name: sequenceAlignment
 //input: string alignType {choices: ['Local alignment', 'Global alignment']}
-// eslint-disable-next-line max-len
 //input: string alignTable {choices: ['AUTO', 'NUCLEOTIDES', 'BLOSUM45', 'BLOSUM50', 'BLOSUM62','BLOSUM80','BLOSUM90','PAM30','PAM70','PAM250','SCHNEIDER','TRANS']}
 //input: double gap
 //input: string seq1
@@ -462,6 +461,11 @@ export async function activityCliffs(table: DG.DataFrame, molecules: DG.Column<s
   similarity: number, methodName: DimReductionMethods,
   similarityMetric: MmDistanceFunctionsNames | BitArrayMetrics, preprocessingFunction: DG.Func,
   options?: (IUMAPOptions | ITSNEOptions) & Options, demo?: boolean): Promise<DG.Viewer | undefined> {
+  //workaround for functions which add viewers to tableView (can be run only on active table view)
+  if (table.name !== grok.shell.tv.dataFrame.name) {
+    grok.shell.error(`Table ${table.name} is not an current table view`);
+    return;
+  }
   if (!checkInputColumnUI(molecules, 'Activity Cliffs'))
     return;
   const axesNames = getEmbeddingColsNames(table);
@@ -588,6 +592,11 @@ export async function sequenceSpaceTopMenu(table: DG.DataFrame, molecules: DG.Co
   plotEmbeddings: boolean, preprocessingFunction?: DG.Func, options?: (IUMAPOptions | ITSNEOptions) & Options,
   clusterEmbeddings?: boolean, isDemo?: boolean
 ): Promise<DG.ScatterPlotViewer | undefined> {
+  //workaround for functions which add viewers to tableView (can be run only on active table view)
+  if (table.name !== grok.shell.tv.dataFrame.name) {
+    grok.shell.error(`Table ${table.name} is not an current table view`);
+    return;
+  }
   const tableView =
     grok.shell.tv.dataFrame == table ? grok.shell.tv : undefined;
   if (!checkInputColumnUI(molecules, 'Sequence Space'))
@@ -965,12 +974,12 @@ export async function manageMonomersView() {
   await monomerManager.getViewRoot();
 }
 
-//name: Monomers
+//name: Manage Monomer Libraries
 //tags: app
 //meta.browsePath: Peptides
 //meta.icon: files/icons/monomers.png
 //output: view v
-export async function manageLibrariesApp(): Promise<DG.View> {
+export async function manageMonomerLibrariesView(): Promise<DG.View> {
   return await showManageLibrariesView(false);
 }
 

package/src/utils/seq-helper/seq-handler.ts

@@ -20,8 +20,6 @@ import {ConvertFunc, ISeqHandler, JoinerFunc, SeqTemps, SeqValueBase} from '@dat
 
 import {SeqHelper} from './seq-helper';
 
-/* eslint-enable max-len */
-
 /** Class for handling notation units in Macromolecule columns and
  * conversion of notation systems in Macromolecule columns
  */
@@ -322,6 +320,44 @@ export class SeqHandler implements ISeqHandler {
     }
   }
 
+  /// Faster method to get monomers at certain position.
+  /// for canonical sequences in fasta (large proteins/nucleotides)
+  /// will be faster than getSplitted(rowIdx).getCanonical(posIdx)
+  getMonomerAtPosition(rowIdx: number, posIdx: number, canonical: boolean): string {
+    if (this.isCanonicalAlphabet && this.isFasta() && !this.getAlphabetIsMultichar()) {
+      const seq = this.column.get(rowIdx) ?? '';
+      const res = seq[posIdx];
+      return canonical ? (res === GapOriginals[NOTATION.FASTA] ? GAP_SYMBOL : (res ?? GAP_SYMBOL)) : (res ?? '');
+    }
+    const mSeq: ISeqSplitted = this.getSplitted(rowIdx);
+    if (posIdx < 0 || posIdx >= mSeq.length)
+      return this.defaultGapOriginal;
+    return canonical ? mSeq.getCanonical(posIdx) : mSeq.getOriginal(posIdx);
+  }
+
+  getMonomersAtPosition(position: number, canonical: boolean): string[] {
+    const length = this.column.length;
+    const res: string[] = new Array(length).fill(GAP_SYMBOL);
+    if (this.isCanonicalAlphabet && this.isFasta() && !this.getAlphabetIsMultichar()) {
+      const colCategories = this.column.categories;
+      const colIndexes = this.column.getRawData();
+      for (let i = 0; i < length; i++) {
+        const seq = colCategories[colIndexes[i]] ?? '';
+        if (position < seq.length) {
+          const resChar = seq[position];
+          res[i] = canonical ? (resChar === GapOriginals[NOTATION.FASTA] ? GAP_SYMBOL : (resChar ?? GAP_SYMBOL)) : (resChar ?? '');
+        }
+      }
+    } else {
+      for (let i = 0; i < length; i++) {
+        const mSeq: ISeqSplitted = this.getSplitted(i);
+        if (position >= 0 && position < mSeq.length)
+          res[i] = canonical ? mSeq.getCanonical(position) : mSeq.getOriginal(position);
+      }
+    }
+    return res;
+  }
+
   /** Any Macromolecule can be represented on Helm format. The reverse is not always possible. */
   public getValue(rowIdx: number, options?: any): SeqValueBase {
     const seq: string = this.column.get(rowIdx);
@@ -344,6 +380,11 @@ export class SeqHandler implements ISeqHandler {
 
   private _stats: SeqColStats | null = null;
 
+  public get isCanonicalAlphabet(): boolean {
+    const alphabet = this.alphabet;
+    return alphabet === ALPHABET.DNA || alphabet === ALPHABET.RNA || alphabet === ALPHABET.PT;
+  }
+
   public get stats(): SeqColStats {
     if (this._stats === null) {
       const freq: { [m: string]: number } = {};
@@ -374,7 +415,7 @@ export class SeqHandler implements ISeqHandler {
   public get maxLength(): number {
     if (this._maxLength === null) {
       this._maxLength = this.column.length === 0 ? 0 :
-        Math.max(...wu.count(0).take(this.column.length).map((rowIdx) => this.getSplitted(rowIdx).length));
+        wu.count(0).take(this.column.length).map((rowIdx) => this.getSplitted(rowIdx).length).reduce((a, b) => a > b ? a : b, 0);
     }
     return this._maxLength!;
   }
@@ -393,6 +434,8 @@ export class SeqHandler implements ISeqHandler {
 
   public isSeparator(): boolean { return this.notation === NOTATION.SEPARATOR || !!this.separator; }
 
+  public isFastaOrSeparator(): boolean { return this.isFasta() || this.isSeparator(); }
+
   public isHelm(): boolean { return this.notation === NOTATION.HELM; }
 
   public isCustom(): boolean { return this.notation === NOTATION.CUSTOM; }