AI-Driven Machine Learning and CRISPR Guide RNA Optimization for Precision Medicine in Neurofibromatosis

Author(s): Shivi Kumar, Deirdre Richardson, Osama Elzafarany, Ray Mikaelson, Laura Gupta

Neurofibromatosis types 1 and 2 (NF1/NF2) are genetically complex disorders with highly variable clinical outcomes and limited precision tools to guide individualized treatment. Current therapies, such as MEK inhibitors, show inconsistent efficacy across patients. There is a critical need for computational approaches that can stratify mutation severity and optimize genome-editing strategies for targeted intervention. To develop and validate an AI-powered precision medicine framework that integrates machine learning for NF1/NF2 mutation classification and disease severity prediction, and deep-learning– driven CRISPR guide RNA (gRNA) optimization to enhance genomeediting accuracy and reduce off-target effects. This was a cross-sectional computational study using public genomic datasets and supervised machine learning models. Deep learning was applied for gRNA scoring. The study was conducted from 2023–2024 and included the development of a web-based deployment tool for real-time analysis. Data analysis and modeling were performed using publicly available databases and Python-based machine learning environments. A Streamlit-hosted web interface was created for realtime clinical and research use. NF1 and NF2 mutation records were curated from ClinVar, HGMD, and LOVD, including over 500 annotated patientderived mutations. Pathogenicity scores were extracted from REVEL and PolyPhen-2. No live human or animal subjects were involved. The machine learning model achieved 93% accuracy for NF1/NF2 mutation classification and 92% precision in disease severity prediction. The deep-learning CRISPR optimization framework demonstrated 98% on-target specificity and a 72% reduction in predicted off-target activity. A web-based interface processed 500 simulated patient sequences with 97% concordance with known ClinVar. This study presents a scalable, AI-driven framework that accurately classifies neurofibromatosis mutations and optimizes CRISPR gRNA selection. The integration of genomic annotation, deep learning, and real-time interface deployment may significantly enhance personalized gene-editing strategies for NF1/NF2 patients. These tools represent a critical step toward translational precision medicine in hereditary tumor syndromes.