A mechanistic free-energy model explains and predicts sequence- and context-dependent CRISPR-Cas9 activity

Accurate prediction of CRISPR-based gene editing remains challenging since existing models often fail to generalize across experimental and cellular contexts. We introduce CRISPRzip, a mechanistic kinetic model that quantitatively links nucleotide sequence and environmental conditions to target interrogation. The model describes R-loop formation as movement through a sequence-dependent free-energy landscape, combining nearest-neighbor nucleic-acid energetics with protein-mediated contributions inferred from high-throughput binding and cleavage kinetics. Applying CRISPRzip to SpCas9, we predict the activity across diverse DNA targets and guide RNAs, and validate with independent singlemolecule FRET and torque spectroscopy experiments. By explicitly incorporating Cas9 concentration and DNA superhelicity, the framework predicts heterogeneous, context-dependent editing outcomes and rationalizes how physical constraints modulate cleavage dynamics. Our results provide a transferable, physics-based foundation that unifies mechanistic insight and predictive power, enabling robust characterization of CRISPR effectors and prediction of their activity across environmental contexts.