@datagrok/bio 1.5.8 → 1.6.0

package/src/utils/notation-converter.ts

@@ -3,9 +3,9 @@ import {WebLogo} from '@datagrok-libraries/bio/src/viewers/web-logo';
 
 /** enum type to simplify setting "user-friendly" notation if necessary */
 export const enum NOTATION {
-  FASTA = 'fasta',
-  SEPARATOR = 'separator',
-  HELM = 'helm'
+  FASTA = 'FASTA',
+  SEPARATOR = 'SEPARATOR',
+  HELM = 'HELM'
 }
 
 /** Class for handling conversion of notation systems in Macromolecule columns */
@@ -32,14 +32,15 @@ export class NotationConverter {
 
   public toHelm(targetNotation: NOTATION): boolean { return targetNotation === NOTATION.HELM; }
 
-  // TODO: isRna
-  public isRna(): boolean { return this.sourceUnits.toLowerCase().endsWith('nt'); }
+  public isRna(): boolean { return this.sourceUnits.toLowerCase().endsWith('rna'); }
+
+  public isDna(): boolean { return this.sourceUnits.toLowerCase().endsWith('dna'); }
 
   public isPeptide(): boolean { return this.sourceUnits.toLowerCase().endsWith('pt'); }
 
   /** Associate notation types with the corresponding units */
   /**
-   * @return {NOTATION} notation associated with the units type
+   * @return {NOTATION}     Notation associated with the units type
    */
   private determineSourceNotation(): NOTATION {
     if (this.sourceUnits.toLowerCase().startsWith('fasta'))
@@ -51,7 +52,33 @@ export class NotationConverter {
       return NOTATION.HELM;
   }
 
-  // TODO: write doc
+  /**
+   * Determine the separator used in SEPARATOR column
+   *
+   * @return {string} The detected separator
+   */
+  private determineSeparator(): string {
+    //   TODO: figure out how to determine the separator efficiently
+    const col = this.sourceColumn;
+    let i = 0;
+    const re = /[^a-z]/;
+    while (i < col.length) {
+      const molecule = col.get(i);
+      const foundSeparator = molecule.toLowerCase().match(re);
+      if (foundSeparator)
+        return foundSeparator[0];
+      i++;
+    }
+    throw new Error('No separators found');
+  }
+
+  /**
+   * Create a new empty column of the specified notation type and the same
+   * length as sourceColumn
+   *
+   * @param {NOTATION} targetNotation
+   * @return {DG.Column}
+   */
   private getNewColumn(targetNotation: NOTATION): DG.Column {
     const col = this.sourceColumn;
     const len = col.length;
@@ -60,74 +87,199 @@ export class NotationConverter {
     // dummy code
     const newColumn = DG.Column.fromList('string', newColName, new Array(len).fill(''));
     newColumn.semType = 'Macromolecule';
-    const newUnits = this.sourceUnits.replace(this.sourceNotation.toString(), targetNotation.toString());
-    newColumn.setTag(DG.TAGS.UNITS, newUnits);
-    // TODO: determine all the qualifiers (units, ...), perhaps, using detectors
+    newColumn.setTag(
+      DG.TAGS.UNITS,
+      this.sourceUnits.replace(
+        this.sourceNotation.toLowerCase().toString(),
+        targetNotation.toLowerCase().toString()
+      )
+    );
+    // TODO: specify cell renderers for all cases
+    if (this.toFasta(targetNotation)) {
+      newColumn.setTag(
+        DG.TAGS.CELL_RENDERER,
+        'Macromolecule');
+    }
     return newColumn;
   }
 
-  // TODO: write doc
-  private convertFastaToSeparator(separator: string): DG.Column {
-    // TODO: implementation
-    // * specify separator
-    // * fasta gap symbol should NOT be considered '-' only, but set as a parameter
-    // * in fasta every position is a monomer (no multi-char monomers), call splitToMonomers() method
-    // * in the resulting jagged array, every gap symbol is to be replaced by
-    // the empty string, while the monomers, to be separated by the separator
-    // (specified as a parameter)
-    // On splitToMonomers(): /libraries/bio/src/viewers/WebLogo --> getSplitter
-
-    const gapSymbol = '-'; // to be specified as an argument
+  /**
+   * Convert a Macromolecule column from FASTA to SEPARATOR notation
+   *
+   * @param {string} separator  A specific separator to be used
+   * @param {string} gapSymbol  Gap symbol in FASTA, '-' by default
+   * @return {DG.Column}        A new column in SEPARATOR notation
+   */
+  private convertFastaToSeparator(separator: string, gapSymbol: string = '-'): DG.Column {
+    // a function splitting FASTA sequence into an array of monomers:
     const splitterAsFasta = WebLogo.splitterAsFasta;
+
     const newColumn = this.getNewColumn(NOTATION.SEPARATOR);
+    // assign the values to the newly created empty column
     newColumn.init((idx: number) => {
-      const sourcePolymer = this.sourceColumn.get(idx);
-      const monomersArray = splitterAsFasta(sourcePolymer);
-      for (let i = 0; i < monomersArray.length; i++) {
-        if (monomersArray[i] === gapSymbol)
-          monomersArray[i] = '';
+      const fastaPolymer = this.sourceColumn.get(idx);
+      const fastaMonomersArray = splitterAsFasta(fastaPolymer);
+      for (let i = 0; i < fastaMonomersArray.length; i++) {
+        if (fastaMonomersArray[i] === gapSymbol)
+          fastaMonomersArray[i] = '';
       }
-      return monomersArray.join(separator);
+      return fastaMonomersArray.join(separator);
     });
     return newColumn;
   }
 
-  private wrapRnaNucleotideToHelm(monomer: string) {
+  /**
+   * Convert a Macromolecule column from FASTA to HELM
+   *
+   * @param {string} fastaGapSymbol   Optional fasta gap symbol
+   * @param {string} helmGapSymbol    Optional helm gap symbol
+   * @return {DG.Column}              A new column in HELM notation
+   */
+  private convertFastaToHelm(
+    fastaGapSymbol: string = '-',
+    helmGapSymbol: string = '*'
+  ): DG.Column {
+    // a function splitting FASTA sequence into an array of monomers
+    const splitterAsFasta = WebLogo.splitterAsFasta;
 
-  }
+    const prefix = (this.isDna()) ? 'DNA1{' :
+      (this.isRna()) ? 'RNA1{' :
+        (this.isPeptide()) ? 'PEPTIDE1{' :
+          'Unknown'; // this case should be handled as exceptional
+
+    if (prefix === 'Unknown')
+      throw new Error('Neither peptide, nor nucleotide');
+
+    const postfix = '}$$$';
+    const leftWrapper = (this.isDna()) ? 'D(' :
+      (this.isRna()) ? 'R(' : ''; // no wrapper for peptides
+    const rightWrapper = (this.isDna() || this.isRna()) ? ')P' : ''; // no wrapper for peptides
 
-  private convertFastaToHelm(): DG.Column {
-    const gapSymbol = '-'; // to be specified as an argument
-    const splitterAsFasta = WebLogo.splitterAsFasta;
     const newColumn = this.getNewColumn(NOTATION.HELM);
+    // assign the values to the empty column
     newColumn.init((idx: number) => {
-      const sourcePolymer = this.sourceColumn.get(idx);
-      const monomersArray = splitterAsFasta(sourcePolymer);
-      for (let i = 0; i < monomersArray.length; i++) {
-        // // TODO: handle gap symbols -- replace by asterisk
-        // if (monomersArray[i] === gapSymbol)
-        //   monomersArray[i] = '*';
-        // else
+      const fastaPolymer = this.sourceColumn.get(idx);
+      const fastaMonomersArray = splitterAsFasta(fastaPolymer);
+      const helmArray = [prefix];
+      let firstIteration = true;
+      for (let i = 0; i < fastaMonomersArray.length; i++) {
+        if (fastaMonomersArray[i] === fastaGapSymbol) {
+          // TODO: verify the correctness of gap symbols handling
+          helmArray.push(helmGapSymbol);
+        } else {
+          const dot = firstIteration ? '' : '.';
+          const item = [dot, leftWrapper, fastaMonomersArray[i], rightWrapper];
+          helmArray.push(item.join(''));
+        }
+        firstIteration = false;
       }
-      // TODO: determine conditionally (if isDna(), or isRna(), or isPeptide()) the template
-      return monomersArray.join('');
+      helmArray.push(postfix);
+      return helmArray.join('');
     });
     return newColumn;
   }
 
-  private convertSeparatorToFasta(): DG.Column {
+  private handleSeparatorItemForFasta(
+    idx: number,
+    separatorItemsArray: string[],
+    separator: string,
+    gapSymbol: string,
+    fastaMonomersArray: string[]
+  ): void {
+    const item = separatorItemsArray[idx];
+    if (item.length > 1) {
+      // the case of a multi-character monomer
+      const monomer = '[' + item + ']';
+      fastaMonomersArray.push(monomer);
+    }
+    if (item === separator) {
+      if (idx !== 0 && separatorItemsArray[idx - 1] === separator)
+        fastaMonomersArray.push(gapSymbol);
+    }
+  }
+
+  private convertSeparatorToFasta(
+    separator: string | null = null,
+    gapSymbol: string = '-'
+  ): DG.Column {
     // TODO: implementation
     // * similarly to fasta2separator, divide string into monomers
     // * adjacent separators is a gap (symbol to be specified)
     // * the monomers MUST be single-character onles, otherwise forbid
+    // * NO, they can be multi-characters
     // conversion
-    //getSplitterWithSeparator
-    return this.getNewColumn(NOTATION.FASTA);
+    // * consider automatic determining the separator
+
+    // if (separator === null)
+    //   separator = this.determineSeparator();
+
+    // a function splitting FASTA sequence into an array of monomers
+    //const splitterAsSeparator = WebLogo.getSplitterWithSeparator(separator);
+    const splitter = WebLogo.getSplitterForColumn(this._sourceColumn);
+
+    const newColumn = this.getNewColumn(NOTATION.FASTA);
+    // assign the values to the empty column
+    newColumn.init((idx: number) => {
+      const separatorPolymer = this.sourceColumn.get(idx);
+      // items can be monomers or separators
+      const separatorItemsArray = splitter(separatorPolymer);
+      const fastaMonomersArray: string[] = [];
+      for (let i = 0; i < separatorItemsArray.length; i++) {
+        const item = separatorItemsArray[i];
+        if (item.length === 0) {
+          fastaMonomersArray.push(gapSymbol);
+        } else if (item.length > 1) {
+          // the case of a multi-character monomer
+          const monomer = '[' + item + ']';
+          fastaMonomersArray.push(monomer);
+        } else {
+          fastaMonomersArray.push(item);
+        }
+      }
+      return fastaMonomersArray.join('');
+    });
+    return newColumn;
   }
 
-  private convertSeparatorToHelm(): DG.Column {
-    // TODO: implementation
-    return this.getNewColumn(NOTATION.HELM);
+  private convertSeparatorToHelm(fastaGapSymbol: string = '-', helmGapSymbol: string = '*'): DG.Column {
+    // a function splitting FASTA sequence into an array of monomers
+    const splitter = WebLogo.getSplitterForColumn(this._sourceColumn);
+
+    const prefix = (this.isDna()) ? 'DNA1{' :
+      (this.isRna()) ? 'RNA1{' :
+        (this.isPeptide()) ? 'PEPTIDE1{' :
+          'Unknown'; // this case should be handled as exceptional
+
+    if (prefix === 'Unknown')
+      throw new Error('Neither peptide, nor nucleotide');
+
+    const postfix = '}$$$';
+    const leftWrapper = (this.isDna()) ? 'D(' :
+      (this.isRna()) ? 'R(' : ''; // no wrapper for peptides
+    const rightWrapper = (this.isDna() || this.isRna()) ? ')P' : ''; // no wrapper for peptides
+
+    const newColumn = this.getNewColumn(NOTATION.HELM);
+    // assign the values to the empty column
+    newColumn.init((idx: number) => {
+      const fastaPolymer = this.sourceColumn.get(idx);
+      const fastaMonomersArray = splitter(fastaPolymer);
+      const helmArray = [prefix];
+      let firstIteration = true;
+      for (let i = 0; i < fastaMonomersArray.length; i++) {
+        if (fastaMonomersArray[i] === fastaGapSymbol) {
+          // TODO: verify the correctness of gap symbols handling
+          helmArray.push(helmGapSymbol);
+        } else {
+          const dot = firstIteration ? '' : '.';
+          const item = [dot, leftWrapper, fastaMonomersArray[i], rightWrapper];
+          helmArray.push(item.join(''));
+        }
+        firstIteration = false;
+      }
+      helmArray.push(postfix);
+      return helmArray.join('');
+    });
+    return newColumn;
   }
 
   private convertHelmToFasta(): DG.Column {
@@ -140,17 +292,25 @@ export class NotationConverter {
     return this.getNewColumn(NOTATION.SEPARATOR);
   }
 
-  /** Dispatcher method for notation conversion */
-  // TODO: write the bodies of converter methods
-  public convert(targetNotation: NOTATION, separator: string | null): DG.Column {
+  /** Dispatcher method for notation conversion
+   *
+   * @param {NOTATION} targetNotation   Notation we want to convert to
+   * @param {string | null} tgtSeparator   Possible separator
+   * @return {DG.Column}                Converted column
+   */
+  public convert(targetNotation: NOTATION, tgtSeparator: string | null = null): DG.Column {
+    // possible exceptions
     if (this.sourceNotation === targetNotation)
-      throw new Error('Target notation is not specified');
-    if (this.isFasta() && this.toSeparator(targetNotation))
-      return this.convertFastaToSeparator(separator!); // there is the only place where a separator is needed
+      throw new Error('Target notation is invalid');
+    if (this.toSeparator(targetNotation) && tgtSeparator === null)
+      throw new Error('Target separator is not specified');
+
+    if (this.isFasta() && this.toSeparator(targetNotation) && tgtSeparator !== null)
+      return this.convertFastaToSeparator(tgtSeparator);
     else if (this.isFasta() && this.toHelm(targetNotation))
       return this.convertFastaToHelm();
     else if (this.isSeparator() && this.toFasta(targetNotation))
-      return this.convertSeparatorToFasta();
+      return this.convertSeparatorToFasta(tgtSeparator!);
     else if (this.isSeparator() && this.toHelm(targetNotation))
       return this.convertSeparatorToHelm();
     else if (this.isHelm() && this.toFasta(targetNotation))

package/{test-Bio-34f75e5127b8-c4c5a3dc.html → test-Bio-34f75e5127b8-95c6fae9.html}

@@ -1,4 +1,4 @@
-<html><head><meta charset="utf-8"/><title>Bio Test Report. Datagrok version datagrok/datagrok:latest SHA=34f75e5127b8. Commit c4c5a3dc.</title><style type="text/css">html,
+<html><head><meta charset="utf-8"/><title>Bio Test Report. Datagrok version datagrok/datagrok:latest SHA=34f75e5127b8. Commit 95c6fae9.</title><style type="text/css">html,
 body {
   font-family: Arial, Helvetica, sans-serif;
   font-size: 1rem;
@@ -229,14 +229,21 @@ 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 c4c5a3dc.</h1></header><div id="metadata-container"><div id="timestamp">Started: 2022-07-11 10:13:14</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">94.078s</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">84.418s</div></div><div class="failureMessages"> <pre class="failureMsg">Error: expect(received).toBe(expected) // Object.is equality
-
-Expected: false
-Received: true
-    at /home/runner/work/public/public/packages/Bio/src/__jest__/remote.test.ts:49:20
-    at Generator.next (&lt;anonymous&gt;)
-    at fulfilled (/home/runner/work/public/public/packages/Bio/src/__jest__/remote.test.ts:31:58)
-    at processTicksAndRejections (internal/process/task_queues.js:97:5)</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/Bio/src/__jest__/test-node.ts:63:11)
+</style></head><body><div id="jesthtml-content"><header><h1 id="title">Bio Test Report. Datagrok version datagrok/datagrok:latest SHA=34f75e5127b8. Commit 95c6fae9.</h1></header><div id="metadata-container"><div id="timestamp">Started: 2022-07-12 11:55:44</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">112.789s</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">100.002s</div></div><div class="failureMessages"> <pre class="failureMsg">Error: thrown: "Exceeded timeout of 100000 ms for a test.
+Use jest.setTimeout(newTimeout) to increase the timeout value, if this is a long-running test."
+    at Object.&lt;anonymous&gt; (/home/runner/work/public/public/packages/Bio/src/__jest__/remote.test.ts:22:1)
+    at Runtime._execModule (/home/runner/work/public/public/packages/Bio/node_modules/jest-runtime/build/index.js:1646:24)
+    at Runtime._loadModule (/home/runner/work/public/public/packages/Bio/node_modules/jest-runtime/build/index.js:1185:12)
+    at Runtime.requireModule (/home/runner/work/public/public/packages/Bio/node_modules/jest-runtime/build/index.js:1009:12)
+    at jestAdapter (/home/runner/work/public/public/packages/Bio/node_modules/jest-circus/build/legacy-code-todo-rewrite/jestAdapter.js:79:13)
+    at runTestInternal (/home/runner/work/public/public/packages/Bio/node_modules/jest-runner/build/runTest.js:389:16)
+    at runTest (/home/runner/work/public/public/packages/Bio/node_modules/jest-runner/build/runTest.js:475:34)
+    at TestRunner.runTests (/home/runner/work/public/public/packages/Bio/node_modules/jest-runner/build/index.js:101:12)
+    at TestScheduler.scheduleTests (/home/runner/work/public/public/packages/Bio/node_modules/@jest/core/build/TestScheduler.js:333:13)
+    at runJest (/home/runner/work/public/public/packages/Bio/node_modules/@jest/core/build/runJest.js:404:19)
+    at _run10000 (/home/runner/work/public/public/packages/Bio/node_modules/@jest/core/build/cli/index.js:320:7)
+    at runCLI (/home/runner/work/public/public/packages/Bio/node_modules/@jest/core/build/cli/index.js:173:3)
+    at Object.run (/home/runner/work/public/public/packages/Bio/node_modules/jest-cli/build/cli/index.js:155:37)</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/Bio/src/__jest__/test-node.ts:63:11)
     at Generator.next (&lt;anonymous&gt;)
     at fulfilled (/home/runner/work/public/public/packages/Bio/src/__jest__/test-node.ts:28:58)
     at processTicksAndRejections (internal/process/task_queues.js:97:5)</pre><pre class="suite-consolelog-item-message">Using web root: http://localhost:8080</pre></div><div class="suite-consolelog-item"><pre class="suite-consolelog-item-origin">    at /home/runner/work/public/public/packages/Bio/src/__jest__/remote.test.ts:24:11
@@ -246,14 +253,4 @@ Received: true
     at Object.&lt;anonymous&gt;.__awaiter (/home/runner/work/public/public/packages/Bio/src/__jest__/remote.test.ts:30:12)
     at Object.&lt;anonymous&gt; (/home/runner/work/public/public/packages/Bio/src/__jest__/remote.test.ts:22:23)
     at Promise.then.completed (/home/runner/work/public/public/packages/Bio/node_modules/jest-circus/build/utils.js:391:28)
-    at new Promise (&lt;anonymous&gt;)</pre><pre class="suite-consolelog-item-message">Testing Bio package</pre></div><div class="suite-consolelog-item"><pre class="suite-consolelog-item-origin">    at /home/runner/work/public/public/packages/Bio/src/__jest__/remote.test.ts:47:11
-    at Generator.next (&lt;anonymous&gt;)
-    at fulfilled (/home/runner/work/public/public/packages/Bio/src/__jest__/remote.test.ts:31:58)
-    at processTicksAndRejections (internal/process/task_queues.js:97:5)</pre><pre class="suite-consolelog-item-message">MSA.is_correct: TypeError: Cannot read properties of undefined (reading 'split')
-sequenceSpace.sequenceSpaceOpens: TypeError: Cannot read properties of undefined (reading 'col')
-sequenceSpace.init: Operation caused an exception (FileSystemException: Cannot open file, path = '/home/grok/data/prod/packages/data/Bio/sample_FASTA.csv' (OS Error: No such file or directory, errno = 2))
-activityCliffs.activityCliffsOpen: TypeError: Cannot read properties of undefined (reading 'columns')
-activityCliffs.init: TypeError: Cannot read properties of undefined (reading 'close')
-activityCliffs.init: Operation caused an exception (FileSystemException: Cannot open file, path = '/home/grok/data/prod/packages/data/Bio/sample_MSA.csv' (OS Error: No such file or directory, errno = 2))
-renderers.afterMsa: Error: Expected "Macromolecule", got "null"
-</pre></div></div></div></div></body></html>
+    at new Promise (&lt;anonymous&gt;)</pre><pre class="suite-consolelog-item-message">Testing Bio package</pre></div></div></div></div></body></html>