Cas12a-Capture: A Novel, Low-Cost, and Scalable Method for Targeted Sequencing

Abstract

Targeted sequencing remains a valuable technique for clinical and research applications. However, many existing technologies suffer from pervasive guanine-cytosine (GC) sequence content bias, high input DNA requirements, and high cost for custom panels. We have developed Cas12a-Capture, a low-cost and highly scalable method for targeted sequencing. The method utilizes preprogrammed guide RNAs to direct CRISPR-Cas12a cleavage of double-stranded DNA in vitro and then takes advantage of the resulting four to five nucleotide overhangs for selective ligation with a custom sequencing adapter. Addition of a second sequencing adapter and enrichment for ligation products generates a targeted sequence library. We first performed a pilot experiment with 7176 guides targeting 3.5 Mb of DNA. Using these data, we modeled the sequence determinants of Cas12a-Capture efficiency, then designed an optimized set of 11,438 guides targeting 3.0 Mb. The optimized guide set achieves an average 64-fold enrichment of targeted regions with minimal GC bias. Cas12a-Capture variant calls had strong concordance with Illumina Platinum Genome calls, especially for single nucleotide variants, which could be improved by applying basic variant quality heuristics. We believe Cas12a-Capture has a wide variety of potential clinical and research applications and is amendable for selective enrichment for any double-stranded DNA template or genome.

FIG. 1.

Schematic outlining the Cas12a-Capture protocol. Genomic DNA (black bars) is dephosphorylated and then treated with Cas12a as well as a pool of gRNAs that cleave target sites (colored regions on genomic DNA) to leave overhangs. A custom biotinylated adapter (beige bars) with degenerate overhangs is ligated to the cleaved molecules, then the other adapter (light blue) is added with Tn5 tagmentation. Finally, a streptavidin pulldown isolates library molecules, which are amplified by on-bead PCR (beige and light blue arrows are primers). gRNAs, guide RNAs.

FIG. 2.

Modeling the sequence determinants of Cas12a-Capture performance. (A) Read uniformity for guides in the pilot experiment. Dashed lines indicate a log10 window, within which 49.3% of guides performed. (B) The 20 features in the linear regression model with the largest positive and negative coefficients. (C) Cross-validation of the linear regression model on fully withheld test data. Pearson r = 0.79. (D) Feature coefficients of individual position-specific nucleotides in the DNA target. PAM, protospacer adjacent motif.

FIG. 3.

Performance of the optimized guide set. (A) Read uniformity for guides in the optimized experiment. Dashed lines indicate a log10 window within which 54.0% of guides performed. (B) Per-base read coverage across the full target with downsampled datasets. (C) Boxplots showing coverage of bases within different 100 bp guanine-cytosine (GC) content bins.