TAD-Guard: A 3D-Genome–Aware Risk Model to Avoid Large-Scale Truncations and Rearrangements

Build a computational safety layer that estimates the probability and extent of catastrophic on-target outcomes (deletions, truncations, translocations) for each candidate guide. Features include distance to TAD boundaries/loop anchors, chromatin compaction, replication timing, nearby fragile sites/repeats, and local p53 response elements. Validate with long-read WGS and optical mapping after editing. This model encodes mechanistic insights from DSB toxicity and existing specificity strategies into a 3D genome–aware guide selector, outputting risk-adapted alternatives for high-risk loci. It aims to prevent low-frequency but high-severity events that can derail therapeutic programs, complementing existing off-target scoring with a new dimension of structural safety, thereby improving regulatory confidence and reducing downstream screening burden.

References:

1. Emerging strategies to minimize the off-target effects in CRISPR/Cas9 system. Chengpei Ouyang (2024). International Conference on Biological Engineering and Medical Science.
2. CRISPR-Cas9 genome editing induces megabase-scale chromosomal truncations. G. Cullot, J. Boutin, J. Toutain, F. Prat, P. Pennamen, C. Rooryck, Martin Teichmann, Emilie Rousseau, I. Lamrissi‐Garcia, V. Guyonnet-Duperat, Alice Bibeyran, Magalie Lalanne, V. Prouzet-Mauléon, B. Turcq, C. Ged, J. Blouin, E. Richard, S. Dabernat, F. Moreau-Gaudry, A. Bedel (2019). Nature Communications.