@aleph-ai/tinyaleph 1.2.1 → 1.3.0

package/backends/bioinformatics/transcription.js

@@ -0,0 +1,221 @@
+/**
+ * Transcription Operator - DNA → RNA
+ * 
+ * Implements the first step of the Central Dogma:
+ * DNA is transcribed into messenger RNA.
+ * 
+ * In prime space: T(2) → U(5) substitution
+ */
+
+const { NUCLEOTIDE_PRIMES, PRIME_COMPLEMENTS, PRIME_TO_NUCLEOTIDE } = require('./encoding');
+
+/**
+ * TranscriptionOperator
+ * 
+ * Transforms DNA prime sequences to RNA prime sequences.
+ * Models the biological process of transcription.
+ */
+class TranscriptionOperator {
+  constructor(options = {}) {
+    this.options = {
+      senseStrand: true,  // If true, use sense strand (same as mRNA except T→U)
+      ...options
+    };
+  }
+  
+  /**
+   * Basic transcription: T → U substitution
+   * @param {number[]} dnaPrimes - DNA sequence as prime array
+   * @returns {number[]} RNA sequence as prime array
+   */
+  apply(dnaPrimes) {
+    return dnaPrimes.map(p => p === 2 ? 5 : p);  // T(2) → U(5)
+  }
+  
+  /**
+   * Full transcription with promoter recognition
+   */
+  transcribe(dnaPrimes, options = {}) {
+    const opts = { ...this.options, ...options };
+    
+    // Find promoter sequence (TATA box analog)
+    const promoterPattern = [2, 7, 2, 7];  // TATA in primes
+    const promoterPos = opts.force ? 0 : this.findPattern(dnaPrimes, promoterPattern);
+    
+    if (promoterPos === -1 && !opts.force) {
+      return {
+        success: false,
+        error: 'No promoter sequence found',
+        dnaPrimes
+      };
+    }
+    
+    // Determine which strand to transcribe
+    let template = dnaPrimes;
+    if (!opts.senseStrand) {
+      template = this.complement(dnaPrimes);
+    }
+    
+    // Find transcription start (+1) - typically ~25bp downstream of TATA
+    const startPos = opts.startPos !== undefined ? opts.startPos :
+      (promoterPos !== -1 ? Math.min(promoterPos + 8, dnaPrimes.length) : 0);
+    
+    // Find terminator (poly-A signal analog or end)
+    const endPos = opts.endPos !== undefined ? opts.endPos :
+      this.findTerminator(template, startPos) || template.length;
+    
+    // Transcribe the region
+    const transcribedRegion = template.slice(startPos, endPos);
+    const rnaPrimes = this.apply(transcribedRegion);
+    
+    return {
+      success: true,
+      rna: rnaPrimes,
+      startPosition: startPos,
+      endPosition: endPos,
+      length: rnaPrimes.length,
+      promoterPosition: promoterPos
+    };
+  }
+  
+  /**
+   * Get complementary strand (template strand from sense strand)
+   */
+  complement(primes) {
+    return primes.map(p => PRIME_COMPLEMENTS[p] || p);
+  }
+  
+  /**
+   * Get reverse complement
+   */
+  reverseComplement(primes) {
+    return this.complement(primes).reverse();
+  }
+  
+  /**
+   * Find pattern in sequence
+   */
+  findPattern(sequence, pattern) {
+    for (let i = 0; i <= sequence.length - pattern.length; i++) {
+      let match = true;
+      for (let j = 0; j < pattern.length; j++) {
+        if (sequence[i + j] !== pattern[j]) {
+          match = false;
+          break;
+        }
+      }
+      if (match) return i;
+    }
+    return -1;
+  }
+  
+  /**
+   * Find all occurrences of pattern
+   */
+  findAllPatterns(sequence, pattern) {
+    const positions = [];
+    for (let i = 0; i <= sequence.length - pattern.length; i++) {
+      let match = true;
+      for (let j = 0; j < pattern.length; j++) {
+        if (sequence[i + j] !== pattern[j]) {
+          match = false;
+          break;
+        }
+      }
+      if (match) positions.push(i);
+    }
+    return positions;
+  }
+  
+  /**
+   * Find terminator sequence
+   * Looks for poly-A signal analog (AATAAA → 7,7,2,7,7,7 in primes)
+   */
+  findTerminator(sequence, startFrom = 0) {
+    const polyASignal = [7, 7, 2, 7, 7, 7];  // AATAAA
+    
+    for (let i = startFrom; i <= sequence.length - polyASignal.length; i++) {
+      let match = true;
+      for (let j = 0; j < polyASignal.length; j++) {
+        if (sequence[i + j] !== polyASignal[j]) {
+          match = false;
+          break;
+        }
+      }
+      if (match) return i + polyASignal.length + 20;  // Include poly-A tail region
+    }
+    return null;
+  }
+  
+  /**
+   * Calculate entropy change from transcription
+   * Transcription slightly increases entropy (expansion of state space)
+   */
+  entropyDelta(dnaPrimes) {
+    return 0.01 * dnaPrimes.length;  // Small positive entropy change
+  }
+  
+  /**
+   * Add 5' cap (methylated G)
+   */
+  addFivePrimeCap(rnaPrimes) {
+    // 7-methylguanosine cap represented as modified G prime
+    const cap = [111];  // Using a distinct prime for m7G
+    return [...cap, ...rnaPrimes];
+  }
+  
+  /**
+   * Add 3' poly-A tail
+   */
+  addPolyATail(rnaPrimes, length = 200) {
+    const tail = Array(length).fill(7);  // Poly-A
+    return [...rnaPrimes, ...tail];
+  }
+  
+  /**
+   * Full mRNA processing: cap + splice + poly-A
+   */
+  processmRNA(rnaPrimes, options = {}) {
+    let processed = [...rnaPrimes];
+    
+    // Add 5' cap
+    if (options.addCap !== false) {
+      processed = this.addFivePrimeCap(processed);
+    }
+    
+    // Splice out introns (if splice sites provided)
+    if (options.introns && options.introns.length > 0) {
+      processed = this.splice(processed, options.introns);
+    }
+    
+    // Add poly-A tail
+    if (options.addPolyA !== false) {
+      const tailLength = options.polyALength || 200;
+      processed = this.addPolyATail(processed, tailLength);
+    }
+    
+    return processed;
+  }
+  
+  /**
+   * Splice out introns
+   * @param {number[]} rnaPrimes - Pre-mRNA
+   * @param {Array<{start: number, end: number}>} introns - Intron positions
+   */
+  splice(rnaPrimes, introns) {
+    // Sort introns by start position (descending) to splice from end first
+    const sortedIntrons = [...introns].sort((a, b) => b.start - a.start);
+    
+    let spliced = [...rnaPrimes];
+    for (const intron of sortedIntrons) {
+      spliced = [
+        ...spliced.slice(0, intron.start),
+        ...spliced.slice(intron.end)
+      ];
+    }
+    
+    return spliced;
+  }
+}
+
+module.exports = { TranscriptionOperator };

package/backends/bioinformatics/translation.js

@@ -0,0 +1,264 @@
+/**
+ * Translation Operator - RNA → Protein
+ * 
+ * Implements the second step of the Central Dogma:
+ * mRNA is translated into protein by ribosomes.
+ * 
+ * Key insight: Translation is an entropy-reducing transform
+ * (64 codons → 21 amino acids = significant information compression)
+ */
+
+const { NUCLEOTIDE_PRIMES, AMINO_ACID_PRIMES, PRIME_TO_NUCLEOTIDE } = require('./encoding');
+const { STANDARD_GENETIC_CODE, isStartCodon, isStopCodon, translateCodon } = require('./genetic-code');
+
+/**
+ * TranslationOperator
+ * 
+ * Transforms RNA prime sequences to protein prime sequences.
+ * Models ribosome-mediated translation.
+ */
+class TranslationOperator {
+  constructor(geneticCode = STANDARD_GENETIC_CODE) {
+    this.geneticCode = geneticCode;
+    
+    // Prime-based start codon: AUG → [7, 5, 11]
+    this.startCodonPrimes = [7, 5, 11];
+    
+    // Prime-based stop codons
+    this.stopCodonPrimes = [
+      [5, 7, 7],   // UAA (Ochre)
+      [5, 7, 11],  // UAG (Amber)
+      [5, 11, 7],  // UGA (Opal)
+    ];
+  }
+  
+  /**
+   * Translate RNA primes to protein primes
+   * @param {number[]} rnaPrimes - RNA sequence as prime array
+   * @param {object} options - Translation options
+   * @returns {object} Translation result
+   */
+  apply(rnaPrimes, options = {}) {
+    const aminoAcids = [];
+    
+    // Find start codon
+    const startPos = options.start !== undefined ? options.start : 
+      this.findStartCodon(rnaPrimes);
+    
+    if (startPos === -1 && !options.force) {
+      return {
+        success: false,
+        error: 'No start codon (AUG) found',
+        protein: [],
+        length: 0
+      };
+    }
+    
+    const start = startPos === -1 ? 0 : startPos;
+    
+    // Translate codon by codon
+    let stoppedAt = -1;
+    for (let i = start; i <= rnaPrimes.length - 3; i += 3) {
+      const codon = rnaPrimes.slice(i, i + 3);
+      
+      // Check for stop codon
+      if (this.isStopCodonPrimes(codon)) {
+        stoppedAt = i;
+        break;
+      }
+      
+      const aa = this.translateCodonPrimes(codon);
+      if (aa !== null) {
+        aminoAcids.push(aa);
+      }
+    }
+    
+    return {
+      success: true,
+      protein: aminoAcids,
+      length: aminoAcids.length,
+      startPosition: start,
+      stopPosition: stoppedAt,
+      readingFrame: start % 3
+    };
+  }
+  
+  /**
+   * Translate a single codon (3 primes) to amino acid prime
+   */
+  translateCodonPrimes(codonPrimes) {
+    if (codonPrimes.length !== 3) return null;
+    
+    // Convert primes to nucleotide string
+    const nucMap = { 7: 'A', 5: 'U', 11: 'G', 3: 'C' };
+    const codonStr = codonPrimes.map(p => nucMap[p] || '?').join('');
+    
+    if (codonStr.includes('?')) return null;
+    
+    const aa = this.geneticCode[codonStr];
+    if (!aa || aa === '*') return null;
+    
+    return AMINO_ACID_PRIMES[aa];
+  }
+  
+  /**
+   * Find start codon (AUG) position
+   */
+  findStartCodon(rnaPrimes) {
+    for (let i = 0; i <= rnaPrimes.length - 3; i++) {
+      if (rnaPrimes[i] === 7 &&      // A
+          rnaPrimes[i + 1] === 5 &&  // U
+          rnaPrimes[i + 2] === 11) { // G
+        return i;
+      }
+    }
+    return -1;
+  }
+  
+  /**
+   * Find all start codons (for alternative translation initiation)
+   */
+  findAllStartCodons(rnaPrimes) {
+    const positions = [];
+    for (let i = 0; i <= rnaPrimes.length - 3; i++) {
+      if (rnaPrimes[i] === 7 &&
+          rnaPrimes[i + 1] === 5 &&
+          rnaPrimes[i + 2] === 11) {
+        positions.push(i);
+      }
+    }
+    return positions;
+  }
+  
+  /**
+   * Check if codon primes represent a stop codon
+   */
+  isStopCodonPrimes(codonPrimes) {
+    for (const stop of this.stopCodonPrimes) {
+      if (codonPrimes[0] === stop[0] &&
+          codonPrimes[1] === stop[1] &&
+          codonPrimes[2] === stop[2]) {
+        return true;
+      }
+    }
+    return false;
+  }
+  
+  /**
+   * Calculate entropy reduction from translation
+   * Translation compresses 64 states to 21 states per codon
+   */
+  entropyDelta(rnaPrimes) {
+    const codons = Math.floor(rnaPrimes.length / 3);
+    // Entropy reduction: log2(64) - log2(21) ≈ 1.6 bits per codon
+    return -codons * Math.log2(64 / 21);
+  }
+  
+  /**
+   * Translate all three reading frames
+   */
+  translateAllFrames(rnaPrimes) {
+    return [0, 1, 2].map(frame => ({
+      frame,
+      ...this.apply(rnaPrimes.slice(frame), { force: true })
+    }));
+  }
+  
+  /**
+   * Find longest open reading frame (ORF)
+   */
+  findLongestORF(rnaPrimes) {
+    let longestORF = null;
+    
+    for (let frame = 0; frame < 3; frame++) {
+      const startCodons = this.findAllStartCodons(rnaPrimes.slice(frame));
+      
+      for (const start of startCodons) {
+        const result = this.apply(rnaPrimes.slice(frame), { start });
+        if (!longestORF || result.length > longestORF.length) {
+          longestORF = {
+            ...result,
+            frame,
+            absoluteStart: frame + start
+          };
+        }
+      }
+    }
+    
+    return longestORF;
+  }
+  
+  /**
+   * Calculate molecular weight of protein
+   */
+  calculateMolecularWeight(proteinPrimes) {
+    const massMap = {
+      23: 57.05,   // G
+      29: 71.08,   // A
+      31: 99.13,   // V
+      37: 113.16,  // L
+      41: 113.16,  // I
+      43: 131.20,  // M
+      47: 147.18,  // F
+      53: 186.21,  // W
+      59: 163.18,  // Y
+      61: 87.08,   // S
+      67: 101.11,  // T
+      71: 103.14,  // C
+      73: 114.10,  // N
+      79: 128.13,  // Q
+      83: 128.17,  // K
+      89: 156.19,  // R
+      97: 137.14,  // H
+      101: 115.09, // D
+      103: 129.12, // E
+      107: 97.12,  // P
+    };
+    
+    // Sum residue masses and subtract water for each peptide bond
+    const residueMass = proteinPrimes.reduce((sum, p) => sum + (massMap[p] || 0), 0);
+    const waterLoss = (proteinPrimes.length - 1) * 18.015;
+    
+    return residueMass - waterLoss;
+  }
+  
+  /**
+   * Calculate isoelectric point (pI) estimate
+   */
+  calculateIsoelectricPoint(proteinPrimes) {
+    // Count charged residues
+    const chargedResidues = {
+      positive: 0,  // K, R, H
+      negative: 0,  // D, E
+    };
+    
+    for (const p of proteinPrimes) {
+      if (p === 83 || p === 89 || p === 97) chargedResidues.positive++;
+      if (p === 101 || p === 103) chargedResidues.negative++;
+    }
+    
+    // Simple pI estimate based on charge balance
+    // Real pI calculation requires pKa values and Henderson-Hasselbalch
+    const netCharge = chargedResidues.positive - chargedResidues.negative;
+    
+    if (netCharge > 0) return 8.0 + Math.log(netCharge + 1);
+    if (netCharge < 0) return 6.0 - Math.log(-netCharge + 1);
+    return 7.0;
+  }
+  
+  /**
+   * Check for signal peptide (N-terminal hydrophobic region)
+   */
+  hasSignalPeptide(proteinPrimes, windowSize = 15) {
+    if (proteinPrimes.length < windowSize) return false;
+    
+    // Signal peptides have hydrophobic N-terminus
+    // Small primes (23-43) are hydrophobic
+    const nTerminal = proteinPrimes.slice(0, windowSize);
+    const hydrophobicCount = nTerminal.filter(p => p >= 23 && p <= 43).length;
+    
+    return hydrophobicCount >= windowSize * 0.6;
+  }
+}
+
+module.exports = { TranslationOperator };

package/backends/index.js

@@ -6,10 +6,34 @@ const { Backend } = require('./interface');
 const { SemanticBackend } = require('./semantic');
 const { CryptographicBackend } = require('./cryptographic');
 const { ScientificBackend } = require('./scientific');
+const bioinformatics = require('./bioinformatics');
 
 module.exports = {
   Backend,
   SemanticBackend,
   CryptographicBackend,
-  ScientificBackend
+  ScientificBackend,
+  
+  // Bioinformatics backend
+  BioinformaticsBackend: bioinformatics.BioinformaticsBackend,
+  
+  // Bioinformatics operators
+  TranscriptionOperator: bioinformatics.TranscriptionOperator,
+  TranslationOperator: bioinformatics.TranslationOperator,
+  FoldingTransform: bioinformatics.FoldingTransform,
+  BindingAffinityCalculator: bioinformatics.BindingAffinityCalculator,
+  MolecularDocker: bioinformatics.MolecularDocker,
+  
+  // DNA Computing
+  DNAStrand: bioinformatics.DNAStrand,
+  DNADuplex: bioinformatics.DNADuplex,
+  ANDGate: bioinformatics.ANDGate,
+  ORGate: bioinformatics.ORGate,
+  NOTGate: bioinformatics.NOTGate,
+  NANDGate: bioinformatics.NANDGate,
+  DNACircuit: bioinformatics.DNACircuit,
+  StrandDisplacementReaction: bioinformatics.StrandDisplacementReaction,
+  
+  // Full bioinformatics module
+  bioinformatics
 };