CRIS.py: A Versatile and High-throughput Analysis Program for CRISPR-based Genome Editing

Abstract

CRISPR-Cas9 technology allows the creation of user-defined genomic modifications in cells and whole organisms. However, quantifying editing rates in pools of cells or identifying correctly edited clones is tedious. Targeted next-generation sequencing provides a high-throughput platform for optimizing editing reagents and identifying correctly modified clones, but the large amount of data produced can be difficult to analyze. Here, we present CRIS.py, a simple and highly versatile python-based program which concurrently analyzes next-generation sequencing data for both knock-out and multiple user-specified knock-in modifications from one or many edited samples. Compared to available NGS analysis programs for CRISPR based-editing, CRIS.py has many advantages: (1) the ability to analyze from one to thousands of samples at once, (2) the capacity to check each sample for multiple sequence modifications, including those induced by base-editors, (3) an output in an easily searchable file format enabling users to quickly sort through and identify correctly targeted clones.

Figure 1

Schema of CRIS.py screening method. (a) Workflow for generating and analyzing genome edited cells and mice. Genome editing reagents are introduced into desired cell type/organism. The cells or embryos are cultured and given time for the editing to occur. Genomic DNA (gDNA) is harvested, target regions amplified and indexed, amplicons pooled and sequenced by NGS. In the final step, sequencing results are analyzed with CRIS.py. (b) Flow chart to analyze NGS results using CRIS.py. Two short user-defined sequences (seq_start and seq_end) are matched to the specified wild-type amplicon sequence (ref_seq) and the total length of characters between the sequences is measured to calculate the wild-type length. (1 and 2) Every sequence in each fastq file is checked for seq_start and seq_end (a match is termed a seq_match), and (3) the length is measured and compared to the reference. (4) all lengths in a fastq file are quantified and reported. (5) Each seq_match is checked for user-defined test_sequences.

Figure 2

Analysis of NHEJ activity in pools of cells. (a) Layout of target locus with gRNAs (g10 and g14), seq_start, and seq_end labeled. (b) A representative CRIS.py output CSV file for CRISPR activity in different pools of cells (fastq files). (c) A representative CRIS.py TXT output file lists the most common reads found in each fastq file along with the read counts for each read with a comma separating the sequence read and the read count. Note that the top reads for the negative control are all of WT length and only vary by single base pair sequencing errors (red base pairs). Treated wells (g10 and g14) contain a large portion of indels as can easily be seen with the aligned reads of different lengths and corresponding read counts. Blue arrows indicated the respective gRNA cut site.

Figure 3

Analysis of CRISPR activity and gene targeting in a pool of cells transfected with gRNA, Cas9, and 2 different ssODNs. (a) Layout of targeted genomic region with the gRNA target site, seq_start, and seq_end labeled. Target modifications are shown in red. Two ssODNs (block_mod_ssODN and block_ssODN) used for gene targeting are aligned to WT genomic locus. (b) The CRIS.py summary CSV file shows cells treated with Cas9, gRNA, and ssODNs contain a high frequency of indels (52.8%). Additionally, 13.4% and 10.4% of the pool underwent correction with the block_mod_ssODN or the block_ssODN, respectively (yellow highlight). (c) CRIS.py TXT file results shows top reads found in each of the 2 pools of cells. Reads resulting from gene targeting by ssODNs highlighted in yellow. Sequencing errors, mod, and block modifications are indicated in red. Blue arrow indicates the gRNA cut site.

Figure 4

CRIS.py analysis of single-cell-derived clones from a pool of cells treated with gRNA, Cas9, and 2 ssODNs. (a) The CRIS.py summary CSV file is easily sortable and allows for quick identification of clones with all combinations of potential modifications. The most relevant columns for each genotype listed are highlighted with corresponding colors. (b) CRIS.py summary TXT file showing most common reads found in 3 representative wells/clones (i., ii., iii.) with a mix of WT, modified, and KO alleles. Sequencing errors, mod, and block modifications highlighted in red. Layout of target locus, gRNAs, and ssODNS is the same as shown in Fig. 3. Blue arrow indicates the gRNA cut site.

Figure 5

CRIS.py analysis of pups injected with gRNA, Cas9, and loxP containing ssODN. (a) Input and Test_sequences for CRIS.py program to analyze pups after direct embryo injection. (b) CRISP.py master summary file results from screened pups. LoxP site integration is detected by both the “loxP” test sequence and in the top identified indels as the loxP integration creates a 40 bp insertion (34 bp for LoxP site + BamHI site).