"Standard models are ranking tools; the Semiprime λ Pipeline is a decision engine for undruggable genomes."
What We Do: Identify viable CRISPR targets in repetitive genomic regions (CAG repeats, low-complexity sequences) that standard tools systematically exclude.
Why It Matters: Billions in R&D are stalled because computational tools treat 20-30% of the genome as "design-dead zones," eliminating therapeutic opportunities for repeat-expansion diseases.
Current gRNA design tools use heuristic rules to filter out repetitive sequences, creating therapeutic blind spots for diseases like Huntington's, Fragile X, and C9orf72 ALS.
We treat DNA as integer compositions and measure algebraic stability via semiprime factorization. Our λ score captures a hidden structural feature of the sequence, allowing us to find viable targets where others cannot.
Hypothesis: Cas9 interacts with sequence context, not just identity. Our scoring captures this hidden layer.
Our model demonstrates a high-precision "NO-GO" filter (the "Left Wall" at λ=0), eliminating non-viable candidates pre-experimentally and proving it is an orthogonal signal, not a proxy for GC content.
We identified viable gRNA candidates within the pathogenic HTT CAG repeat tract—a "no-fly zone" for standard tools, creating novel IP in uncontested white space.
- ✅ Completed: Computational validation on 11K+ benchmark sequences.
- 🔄 In Progress: Provisional patent filing & wet-lab partner outreach.
- 📈 Roadmap: Validate HTT candidates -> Scale to 3-5 other repeat diseases -> Launch SaaS platform.
This is a platform play with a portfolio of targets, not a single high-risk asset.
git clone https://github.com/Joedaddy66/integer-resonance-crispr.git
cd integer-resonance-crispr
pip install -r requirements.txt
# Download full Doench 2016 dataset and place in root
# python analyze.pyWet Lab Partners, Strategic Investors, and Pharma Partnerships.
Contact: [JPurvis6691@gmail.com]