2026· BIO Web of Conferences· Vol 240, pp. 01003· 0 citations· 2 references
TL;DR
Major bioinformatics tools for CRISPR are reviewed, identifying the layered separation as a key bottleneck and highlighting the demand for a unified programming framework to streamline workflows and advance CRISPR algorithm innovation.
Abstract
CRISPR-Cas9 has revolutionized molecular biology with its precise genome-editing ability, whose precision largely relies on computationally designed guide RNAs (gRNAs) to minimize off-target effects. This paper reviews major bioinformatics tools for CRISPR, focusing on their algorithms and computational frameworks, classifying them into four layers. The first consists of rule-based systems such as CHOPCHOP and CRISPOR, using linear scoring functions with O(n) complexity. The second includes alignment-based tools like Cas-OFFinder, which applies FM-index for off-target detection with over 90% sensitivity. The third involves machine learning-based methods such as CRISPRscan with random forest, yet suffering 50–70% false positives due to expanded search space. The fourth layer comprises deep learning systems including CRISMER and DeepCRISPR, employing Transformers and CNNs with AUC 0.85–0.95, though they lack cross-dataset generalization. Moreover, current tools are fragmented, requiring manual integration of multiple platforms and poor reproducibility. This review identifies the layered separation as a key bottleneck and highlights the demand for a unified programming framework to streamline workflows and advance CRISPR algorithm innovation.
Abstract Summary CRISPR-Cas9 has become a widely used tool for genome editing. However, its off-target cleavage caused by partial sequence matches with guide RNAs (gRNAs) remains a critical limitation. Recently, abasic gRNAs (ØXØ) have been developed to enhance target specificity, but their effects vary depending on the positional sequence context. Here, we present abCRISPR, a deep neural network (DNN) framework for the rational design of ØXØ sequences with minimized off-target activity. abCRISPR leverages informative few-shot training with paired datasets of abasic and unmodified gRNAs, using high-quality random mismatch target libraries, exhaustively sequenced for mismatched off-target substrates (n = 97583) in in vitro CRISPR-Cas9 cleavage experiments. Predicted off-target activities for both abasic and unmodified gRNAs showed strong correlation with experimental data (r ≥ 0.95, 10-fold cross-validation). Notably, these comprehensive training sets provide robust ground-truth negatives, enabling accurate and sensitive prediction of off-targets. For unmodified gRNAs, abCRISPR (AUC = 0.98) was validated to outperform existing deep learning-based methods (AUC = 0.45–0.68). When applied to the human genome, abCRISPR generated ØXØ sequences, covering 58 875 004 potent CRISPR-targetable sites with improved target specificity. Together, this work provides a comprehensive bioinformatics resource for safe and precise CRISPR-Cas9 genome editing. Availability and implementation The source code for abCRISPR and training data are available at https://doi.org/10.5281/zenodo.20398246. abCRISPR results for the human genome are available at http://clip.korea.ac.kr/abCRISPR/
Geun-Woo D. Kim, Dowoon Gu, Mingyo Park et al.· Bioinformatics· 0 citations
Accurate identification of CRISPR-Cas9 off-target sites is essential for the safety assessment of genome-editing-based therapies. While numerous in silico prediction tools have been developed, their comparative performance and practical utility in preclinical workflows remain incompletely defined. We performed a systematic benchmarking of 14 in silico CRISPR-Cas9 off-target prediction tools, including both standard approaches and machine learning-based models. The analysis was based on a curated dataset derived from the CRISPRoffT database, comprising 3,827 deep-sequenced genomic sites across 26 guide RNA/Cas9 combinations in human cells. Sites with indel frequencies ≥0.1% were operationally defined as true off-targets. We evaluated tool performance using score distributions, correlation with indel frequencies, precision-recall characteristics, recall among top-ranked candidate sites, and the effect of combining tools. All tools assigned higher scores to true off-target sites compared with nontarget sites, although substantial overlap between classes was observed. Correlation between prediction scores and indel frequencies was weak to moderate, indicating limited ability to predict editing magnitude. Precision-recall performance was moderate across all tools, reflecting inherent trade-offs between sensitivity and specificity. Recall increased with the number of predicted sites considered, reaching approximately 77% among the top 500 and up to 83% among the top 1,250 sites, but leaving a substantial fraction of true off-targets undetected. Combining tools yielded only modest improvements. Current in silico tools enable prioritization of CRISPR-Cas9 off-target candidates but remain limited in their ability to comprehensively identify and quantitatively predict off-target activity. Our findings highlight the importance of considering both ranking performance and candidate site coverage and support the use of combined computational and experimental strategies for robust off-target assessment in preclinical gene editing workflows.
M. M. Kaufmann, Maren Hackenberg, William Jobson Pargeter et al.· Human Gene Therapy· 0 citations
Off-target cleavage is a central safety concern for CRISPR-Cas9 genome editing, particularly in therapeutic applications where unintended double-strand breaks carry clinical risk. We benchmarked five machine learning classifiers — logistic regression on mismatch-count summary features, a random forest and a gradient boosting model on one-hot-encoded sgRNA/candidate-site sequence pairs, a one-dimensional convolutional neural network (CNN) over the positional mismatch map, and a gradient-boosting/CNN ensemble — on a real, published GUIDE-seq off-target dataset (Kleinstiver et al., 2016, Nature) comprising 95,829 candidate off-target sites for five sgRNAs, of which only 54 (0.06%) were experimentally validated as true cleavage sites. On a held-out, stratified test split (n = 19,166; 11 true positives), gradient boosting on combined mismatch and sequence features performed best (ROC-AUC = 0.997, PR-AUC = 0.355, best F1 = 0.50), outperforming a random forest on raw sequence encoding alone (PR-AUC = 0.083) and a sequence CNN (PR-AUC = 0.129). Because the positive class is extremely rare, we report precision-recall AUC as the primary metric rather than ROC-AUC, which is inflated by the large negative class. A positional mismatch analysis showed that experimentally validated off-target sites carried substantially fewer mismatches overall than non-cleaved candidate sites (mean 3.6 vs. 5.9 mismatches across the 23-nucleotide target), and were markedly more mismatch-intolerant in the 10-nucleotide PAM-proximal seed region (11.3% vs. 27.4% per-position mismatch rate) and at the PAM itself (6.8% vs. 16.0%), consistent with established seed-region and PAM-sensitivity models of Cas9 target recognition. We report these findings, including the low absolute precision achievable in this severely imbalanced, small-positive-class setting, as a realistic picture of what off-target classifiers can and cannot yet deliver from sequence alone.
Accurate identification and classification of CRISPR-Cas systems are crucial for understanding microbial immune mechanisms and developing novel genome-editing tools. However, traditional homology-based mining methods face severe computational bottlenecks and assembly fragmentation challenges when processing massive metagenomic data. Here, we present DeepCRISPR-Typer, a comprehensive computational framework integrating a large protein language model (TEMC-Cas), a deep sequence feature extractor (CRISPR-RepTyper), and an adaptive targeted HMM profiling strategy. DeepCRISPR-Typer integrates array and Cas evidence and significantly reduces computational overhead by dynamically invoking subtype-specific HMM subsets. In metagenomic dataset evaluations, DeepCRISPR-Typer achieved a classification accuracy of 94.26% and demonstrated a significant acceleration of approximately 1 orders of magnitude compared to existing mainstream tools. This research provides a robust and scalable engine for metagenome-scale CRISPR system discovery, significantly expanding the mining toolbox for genome engineering applications.
Long Wen, Minghui Jing, Yanyan Li et al.· ACS Synthetic Biology· 0 citations
Abstract Genome editing researchers currently navigate multiple disconnected tools for guide RNA (gRNA) design, primer generation, and editing analysis—a fragmented workflow that introduces errors and limits reproducibility. CasPINS (Cas-Primer-Indel Suite) addresses this gap as an open-source, unified platform integrating the complete genome editing computational workflow into a single interactive web application accessible without programming expertise. The platform supports 90+ species, 14 CRISPR-Cas variants, TALEN design, and six editing modes. Primer design integrates with Ensembl and NCBI databases relative to cut sites, while indel quantification utilizes Non-Negative Least Squares (NNLS) decomposition of Sanger chromatograms with maximum signal extraction and R2-corrected conservative modes. Benchmarking demonstrates strong concordance with established tools, including a 68.8% recovery of CHOPCHOP gRNAs and 67.2% of CRISPOR gRNAs across five human benchmark genes, alongside an algorithmic agreement within 2.6 percentage points on gold-standard TIDE data. Ultimately, a step-count analysis shows that CasPINS significantly streamlines usability, reducing discrete user actions from 25 to 8 steps compared to the traditional sequential-tool pipeline.
R. Dasgupta, Kaushik Das· Bioinformatics Advances· 0 citations
Multidomain proteins play central roles in cellular regulation, yet their intrinsic flexibility and structural instability often hinder optimization for biotechnological applications. Here, we present an integrated structure-guided and deep learning-assisted engineering strategy that combines structure modeling with Protein Message Passing Neural Network (ProteinMPNN)-based sequence design to generate an ultracompact CRISPR activator (uCRISPRa) derived from the miniature CRISPR/Cas12f. Structural and computational analyses identified flexible, nonessential regions within both Cas12f and its single-guide RNA (sgRNA), enabling rational truncation and sequence redesign while preserving DNA-targeting capability. When delivered as mRNA encapsulated in lipid nanoparticles, uCRISPRa achieved selective activation of olfr544 among more than a thousand homologous olfactory receptor genes in skeletal muscle cells, leading to enhanced mitochondrial biogenesis. These findings demonstrate that the integration of structure-based protein engineering with deep learning sequence optimization provides a powerful framework for developing compact and efficient CRISPR effectors, offering broad potential for precise gene regulation and functional studies of complex macromolecular systems.
Smith Le, Trung Thach· The FEBS Journal· 0 citations