GUIDES

GUIDES designs custom CRISPR-Cas9 knockout libraries targeting user-selected gene sets in the human and mouse genomes to maximize on-target efficiency and minimize predicted off-target cleavage.

Key Features:

• Input formats: Accepts gene symbols or standard identifiers as input for library design.
• CCDS-based coding sequences: Obtains coding sequences from CCDS annotations.
• GTEx expression prioritization: For human genes, uses GTEx tissue-specific expression to prioritize constitutively expressed exons and reduce ineffective edits in alternatively spliced exons.
• Genome-wide off-target prediction: Evaluates candidate Cas9 target sites using cutting frequency determination (CFD) scoring for off-target risk.
• On-target efficiency modeling: Estimates on-target cleavage efficiency with boosted regression tree models trained on saturation mutagenesis screens.
• Protein-domain targeting: Preferentially selects sgRNAs that target Pfam-annotated protein domains to increase likelihood of functional disruption.
• Sequence filters: Filters out homopolymer sequences, Pol III terminators, and targets in risky terminal exons.
• Controls: Supports optional inclusion of non-targeting control sgRNAs.
• Oligo output: Generates synthesis-ready oligos with flanking primers for cloning into pooled libraries.
• Benchmarking: Benchmark analyses across 77 pooled screens report stronger depletion of essential genes for GUIDES-designed sgRNAs versus matched controls.
• Performance: Algorithmic implementation exhibits linear runtime relative to gene count (for example, ~15 s for 500 genes × 6 sgRNAs on standard hardware).
• Data integration: Integrates multiple orthogonal data sources and predictive models in a multi-stage pipeline.

Scientific Applications:

• Focused CRISPR screening: Design of high-coverage, custom CRISPR-Cas9 knockout libraries for pooled screens in human and mouse.
• Tissue-specific knockout prioritization: Prioritization of exons constitutively expressed in relevant tissues to improve knockout effectiveness in the presence of alternative splicing.
• Functional disruption of proteins: Targeting of Pfam-annotated domains to increase likelihood of disrupting protein function.
• Pooled library synthesis and screening: Generation of synthesis-ready oligo pools and controls for cloning and pooled functional screens, including identification of essential genes.

Methodology:

Multi-stage pipeline that obtains coding sequences from CCDS, uses GTEx expression to prioritize human exons, evaluates candidate Cas9 sites with genome-wide off-target CFD scoring, estimates on-target efficiency with boosted regression tree models trained on saturation mutagenesis screens, preferentially selects sgRNAs targeting Pfam domains, filters homopolymers/Pol III terminators/terminal-exon targets, and outputs synthesis-ready oligos with flanking primers.