Back

Features affecting Cas9-Induced Editing Efficiency and Patterns in Tomato: Evidence from a Large CRISPR Dataset

Cucuy, A.; Ben-Tov, D.; Melamed-Bessudo, C.; Honig, A.; Cohen, B. A.; Levy, A. A.

2026-01-07 plant biology
10.64898/2026.01.06.696182 bioRxiv
Show abstract

CRISPR/Cas9 is a cornerstone of genome editing, yet the determinants of editing efficiency and DNA Double Strand Break (DSB) repair outcomes remain poorly understood, particularly in plants. To address this gap, we generated a dataset of 420 sgRNAs targeting promoters, exons, and introns of 137 genes in tomato protoplasts and quantified editing efficiencies together with ATAC-seq-derived chromatin accessibility and transcriptional states in the same cellular system. Editing efficiency was consistently higher at open chromatin sites and modestly elevated in promoters and introns relative to exons, whereas transcriptional activity did not measurably influence editing outcomes. Additionally, we identified a local genomic effect resulting in less variable editing between sgRNAs targeting the same compared to different genes. A distinct subset of sgRNAs achieved nearly complete editing, producing long deletions with extended microhomology tracts. These repair footprints closely parallel those observed for high-efficiency guides in human datasets, implicating conserved sequence-driven biases and a predominant role for microhomology-mediated end joining at these sites. Yet, widely used human-trained prediction models failed to rank sgRNA performance in plants, underscoring the limits of cross-species generalization. This dataset defines how chromatin accessibility, genomic context, and intrinsic sequence characteristics shape Cas9 activity in plants, and provides a resource for improving guide design and advancing mechanistic understanding of plant DNA repair.

Matching journals

The top 5 journals account for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.