CasFinder: Flexible algorithm for identifying specific Cas9 targets in genomes

Abstract

CRISPR/Cas9 systems enable many molecular activities to be efficiently directed in vivo to user-specifiable DNA sequences of interest, including generation of dsDNA cuts and nicks, transcriptional activation and repression, and fluorescence. CRISPR targeting relies on base pairing of short RNA transcripts with their target DNA sequences that must also be adjacent to fixed DNA motifs. However, rules for Cas9 targeting specificity are incompletely known. With increasing numbers of Cas9 systems being developed and deployed in more and more organisms, there is now strong need for a flexible and rational method for finding Cas9 sites with low off-targeting potential. We address this through the CasFinder system, which we demonstrate by generating human and mouse exome-wide catalogs of specific sites for three varieties of Cas9 - S. pyogenes, S. thermophilus (ST1), and N. meningitidis - that each target 56-74% of all exons. We also generate reduced sets of up to 3 targets per gene for use in high-throughput Cas9-based gene knockout screens that target 75-80% of all genes.