Applications of the method and development of the protocol

A variety of computational tools have been developed for the design and analysis of CRISPR-based experiments. However, these tools are typically developed in isolation, without features for facile integration with one another, and/or without sufficient consideration to facilitate implementation in a laboratory setting. Here, we offer a protocol to integrate robust, publicly available tools for the design, execution, and analysis of CRISPR genome-editing experiments. Specifically, we have adapted CRISPOR⁹ and CRISPResso¹⁰ to be integrated with one another as well as streamlined for experimental implementation (Figs. 1–3). This protocol has been used in previously published works to functionally interrogate the BCL11A enhancer as well as to evaluate potential functional sequences within all DNase hypersensitive sites (DHSs) in proximity to the MYB gene^11,12. CRISPR mutagenesis allows for the study of both coding and noncoding regions of the genome¹³. This involves usage of one sgRNA for arrayed experiments or multiple sgRNAs for pooled experiments. Arrayed experiments are useful when a target can be mutagenized by one sgRNA. Pooled screening allows for targeting of a handful of genes up to genome-scale gene targeting^14–17. It also allows for saturating mutagenesis (tiling of sgRNA) experiments to identify functional sequence within noncoding regions^11,12,18. This protocol can be used to design and execute arrayed or pooled genome-editing experiments. Furthermore, it can also be used for the design, implementation, and analysis of pooled screens for gene targeting, saturating mutagenesis of noncoding elements, or any other targeting strategies^11,12,19,20. It is important to note that computational skills are required for the command-line and Docker versions of CRISPOR and CRISPResso in this protocol. Specifically, users are required to have a basic understanding of how to execute a command in a terminal and how to navigate a Unix-based file system. However, web-based versions are also available that do not require these skills.

Open in a separate window

Figure 1

Schematic of an arrayed genome-editing experiment. Arrayed genome-editing experiments are performed by designing one sgRNA using CRISPOR. This schematic demonstrates the design of an sgRNA to mutagenize a GATA motif. After designing the optimal sgRNA, it is cloned into pLentiGuide-puro, lentivirus is produced, cells are transduced, and successful transductants are selected (successful transduction is indicated by red curved lines). After conclusion of the experiment, cells are pelleted, and genomic DNA is extracted. Locus-specific PCR primers are used to amplify regions flanking the double-strand break site. Deep sequencing of the amplicon generated by locus-specific PCR is subsequently performed. Quantification of editing frequency and indel distribution is determined by CRISPResso. The GATA motif is underlined, the triangle indicates the double-strand break position, and the PAM sequence is shown in blue.

Open in a separate window

Figure 3

Design, experimental execution, and data analysis workflows for arrayed and pooled genome-editing experiments. sgRNA design steps by CRISPOR are shown in gray, experimental execution steps are shown in blue, and data analysis steps by CRISPResso are shown in red/pink.

CRISPOR: sgRNA and PCR primer design for arrayed and pooled screen experiments

CRISPOR (http://crispor.org) is a computational sgRNA design tool that predicts off-target cleavage sites and offers a variety of on-target efficiency scoring systems to assist sgRNA selection for more than 120 genomes using many different CRISPR nucleases (Supplementary Table 1)⁹. CRISPOR offers a variety of on- and off-target prediction scores that can aid in optimal sgRNA selection (Box 1)⁹. Analysis of on-target sgRNA efficiency can be predicted based on available scores and/or investigated experimentally by analysis of editing frequency. In addition to the numerous sgRNA efficiency prediction scores, CRISPOR offers automated primer design to facilitate PCR amplification of regions for deep-sequencing analysis to quantitate editing frequency. This involves PCR amplification of sequences flanking the DSB site for a given sgRNA. Similarly, CRISPOR offers computational prediction of off-target sites as well as PCR primers for deep-sequencing analysis of potential mutagenesis at these predicted off-target sites.

CRISPOR can also be used for the design of gene-targeted pooled screens by including exonic regions for sgRNA design as well as saturating mutagenesis screens. Alternatively, a list of gene names can be input into CRISPOR (‘CRISPOR Batch’) to aid in the design of large scale gene-targeted libraries, which includes non-targeting controls. This analysis includes automated design of the required oligonucleotides for library cloning.

Saturating mutagenesis involves using all PAM-restricted sgRNAs within a given region(s) in a pooled screening format to identify functional sequences^11,12,18. Saturating mutagenesis can be used to analyze coding and noncoding elements in the genome or a combination of the two. Screen resolution is a function of PAM frequency and can be enhanced by PAM choice and/or a combination of nucleases with unique PAM sequences¹². CRISPOR can be used to design saturating mutagenesis libraries by simply selecting all sgRNAs within the inputted region(s). This analysis also includes automated design of the required oligonucleotides for library cloning. It is particularly important to consider off-target prediction scores offered by CRISPOR for saturating mutagenesis screens, as repetitive sequences can confound screen results¹². sgRNAs with high probability of off-target mutagenesis can be excluded at the library design stage or can be appropriately handled at the analysis stage after the experiments have been performed (Box 1).

CRISPResso: analysis of deep sequencing from arrayed or pooled sgRNA experiments

CRISPResso is a computational pipeline that enables accurate quantification and visualization of CRISPR genome-editing outcomes, as well as comprehensive evaluation of effects on coding sequences, noncoding elements, and off-target sites from individual loci, pooled amplicons, and whole-genome deep-sequencing data¹⁰. The CRISPResso suite involves multiple tools for analysis, including the CRISPResso webtool and command-line version of CRISPResso. There are also multiple additional command-line-only tools in the CRISPResso suite: CRISPRessoPooled, CRISPRessoWGS, CRISPRessoCount, CRISPRessoCompare, and CRISPRessoPooledWGSCompare. The applications and features of these tools are summarized in Table 1. CRISPResso analysis offers many unique features, such as splice-site or frameshift analysis to quantify the proportion of engendered mutations that result in a frameshift when targeting coding sequence. In addition, features for indel visualization have been added to CRISPResso since its initial publication¹⁰. Notably, CRISPResso also provides a variety of features to offer users the opportunity to optimize analysis of sequencing data (Box 2).

CRISPResso analysis of an individual locus requires PCR amplification of the sequences flanking the genomic position of the DSB for a given sgRNA. The resulting deep-sequencing FASTQ file can be analyzed by CRISPResso to quantitate the indel spectrum as well as visualize individual alleles (see Fig. 4 and ANTICIPATED RESULTS). This analysis can be performed when targeting coding (Fig. 4a–d) or noncoding sequences. When targeting coding (exonic) sequence, CRISPResso can determine the frequency of in-frame and out-of-frame (frameshift) mutations produced (Fig. 4d). This type of analysis may suggest toxicity upon gene knockout by identifying an increased frequency of in-frame mutations¹². Two separate CRISPResso analyses with the same amplicon can be directly compared using the CRISPRessoCompare tool (Table 1). CRISPRessoCompare is useful in situations such as comparison of ‘treated’ and ‘untreated’ groups, as well as comparison of different experimental conditions. It can also be used to compare indel distributions created by two different sgRNAs within the same region/amplicon¹¹.

Open in a separate window

Figure 4

Locus-specific deep-sequencing analysis of coding and noncoding targeting by CRISPResso. (a) Frequency distribution of alleles with indels (shown in blue) and without indels (shown in pink) for an sgRNA targeting BCL11A exon 2. (b) All reads with sequence modifications (insertions, deletions, and substitutions) are mapped to a position within the BCL11A exon 2 reference amplicon. The vertical dashed line indicates the position of predicted Cas9 cleavage. The position of the sgRNA is shown in gray. (c) Distribution of indel sizes when targeting BCL11A exon 2. Percentage of unmodified sequences is shown in red and percentages of modified sequences are shown in blue. (d) Frameshift analysis of BCL11A exon 2 coding sequence targeted reads. Frameshift mutations are shown in red and in-frame mutations are shown in tan. (e) sgRNA enrichment based on analysis of fetal hemoglobin (HbF) levels when performing saturating mutagenesis of BCL11A exon 2 and analysis of the functional core of the BCL11A enhancer using NGG- and NGA-restricted sgRNAs from two previously published studies11,12. Nontargeting sgRNAs are pseudomapped with 5-bp spacing.

CRISPResso analysis can be extended from a single amplicon to multiple amplicons using the CRISPRessoPooled tool (Table 1). This is useful for individual locus experiments that require multiple amplicons for analysis of the full region or when multiple genes are targeted. The CRISPResso suite can also be used to analyze whole-genome sequencing (WGS) data through CRISPRessoWGS. This requires pre-aligned WGS data in BAM format, which can be created using publicly available aligners (e.g., Bowtie2 (ref. 21) or Burrows–Wheeler Aligner^22,23). Similar to CRISPRessoCompare, two analyses using either CRISPRessoPooled or CRISPRessoWGS can be directly compared using CRISPRessoPooledWGSCompare (Table 1). This can be particularly useful when using multiple-amplicon sequencing data or WGS data to evaluate off-target cleavages by two different sgRNAs to identify sgRNAs with lower off-target activity.

CRISPResso also offers the ability to analyze deep-sequencing data from pooled CRISPR screens. Sequence/indel-based analysis in a pooled screening format is confounded by non-edited reads from cells containing sgRNAs targeting other regions/loci. Therefore, pooled screens are typically analyzed by enumeration of sgRNAs by PCR-amplifying sequences containing the cloned sgRNA within the lentiviral construct that has been integrated into the cell’s genome (using PCR primers specific to the lentiviral construct). Deep-sequencing data generated using PCR primers specific to the lentiviral construct can be analyzed by CRISPRessoCount for sgRNA enumeration for the purposes of calculating enrichment and/or dropout of sgRNAs under different experimental conditions^11,12 (Figs. 2, ​,4e).4e). After sgRNA enumeration, it is important to normalize the reads by taking into account the total number of reads when comparing two different deep-sequencing samples. An example strategy for performing normalization involves normalizing all samples to 1 million reads. This can be accomplished by dividing the read count for each sgRNA by the total number of reads in that sample. Then this quotient is multiplied by 1 million, for example: (sgRNA_n read count/total read count of sample)×1,000,000. This should be repeated for all samples so that each sample has 1 million total normalized reads. These normalized reads can then be used to calculate enrichment and/or dropout (‘depletion’) ratios, for example: log₂(normalized read count of sgRNA_n in sample X/normalized read count of sgRNA_n in sample Y).

Open in a separate window

Figure 2

Schematic of a pooled genome-editing experiment. Pooled genome-editing experiments are performed by designing multiple sgRNA using CRISPOR. After designing the sgRNAs, they are batch-cloned into pLentiGuide-puro, lentivirus is produced, cells are transduced at low multiplicity, and successful transductants are selected (successful transduction is indicated by curved lines). Phenotypic selection is performed (e.g., FACS, drug/toxin resistance, drug/toxin sensitivity, cell lethality/gene essentiality, cellular fitness/proliferation). After conclusion of the experiment, cells are pelleted, and genomic DNA is extracted. PCR primers specific (primer sequence is underlined) to the pLentiGuide-Puro construct are used to amplify regions flanking the cloned sgRNA sequence. Deep sequencing of the amplicon generated by construct-specific PCR is subsequently performed. sgRNAs present within the sample are enumerated by CRISPRessoCount.

Comparison with other methods

Numerous computational tools are freely available to aid sgRNA design for a wide spectrum of PAM sequences, as well as for on-target efficiency and off-target cleavage predictions: Broad GPP Portal²⁴, Cas-Database²⁵, Cas-OFFinder²⁶, CasOT²⁷, CCTop²⁸, COSMID²⁹, CHOPCHOP^30,31, CRISPRdirect³², CRISPR-DO³³, CRISPR-ERA³⁴, CRISPR-P³⁵, CROP-IT³⁶, DNA Striker¹², E-CRISP³⁷, flyCRISPR³⁸, GUIDES³⁹, GuideScan⁴⁰, GT-scan⁴¹, MIT CRISPR design tool⁸, WU-CRISPR⁴², CRISPRseek⁴³, sgRNAcas9 (ref. 44), and CRISPR multiTargeter⁴⁵, as well as others offered by companies such as Deskgen⁴⁶ and Benchling⁴⁶. CRISPOR offers several unique advantages for designing sgRNAs for genome-editing experiments. First, CRISPOR integrates multiple published on-target sgRNA efficiency scores, including those from Fusi et al.⁴⁷, Chari et al.⁴⁸, Xu et al.⁴⁹, Doench et al.^24,50, Wang et al.⁵¹, Moreno-Mateos et al.⁵², Housden et al.⁵³, Prox. GC⁵⁴, -GG⁵⁵, and Out-of-Frame⁵⁶. It also offers previously published off-target prediction (MIT specificity score)⁸. Second, CRISPOR has been optimized to facilitate experimental implementation by providing automated primer design for both on-target and off-target deep-sequencing analysis. The primers and output files are further designed to be compatible for subsequent analysis by CRISPResso after the experiments have been completed. CRISPOR also offers features to automate oligonucleotide design for cloning pooled sgRNA libraries targeting both genes and noncoding regions. Taken together, CRISPOR provides an sgRNA design platform to facilitate experimental execution and downstream data analysis.

Alternative computational tools to CRISPResso exist to evaluate genome-editing outcomes from deep-sequencing data^57–61; however, these tools offer limited analysis functionality for pooled amplicon sequencing or WGS data as compared with the CRISPResso suite. CrispRVariants is another tool that offers functionality to analyze deep-sequencing data by quantifying mosaicism and allele-specific gene editing, as well as multisequence alignment views⁶⁰.

CRISPR genome-editing reagents have taken many forms, including DNA, RNA, protein, and various combinations of each^62,63. Delivery of these reagents has also been attempted using a variety of methods, including electroporation, lipid-based transfection, and viral-mediated delivery^62,63. For further discussion of delivery methods for genome-editing reagents, refer to Yin et al.⁶⁴. Pooled screening relies on the ability to deliver individual reagents to individual cells in batches⁶⁵. Electroporation and lipid-based transfection methods offer limited ability to control the number of reagents (i.e., sgRNAs) delivered per cell; however, lentiviral transduction at low transduction rates (~30–50%) results in single lentiviral integrants per cell in the majority of cases⁶⁵. Furthermore, lentivirus offers stable integration of the CRISPR reagents into each cell’s genome. These features of lentivirus allow for pooled CRISPR experiments. Therefore, this protocol describes the use of lentivirus for both arrayed and pooled CRISPR experiments (Boxes 3 and 4).

Limitations of on-target prediction, off-target prediction, and indel analysis

As described above, there are many tools available for both on-target sgRNA efficiency and off-target cleavage prediction. Progress has been made toward enhancing the predictive value of these scores; however, although these predictions are useful to focus sgRNA selection for experimental design, experimental validation provides the definitive analysis of on-target and off-target mutagenesis. Alternatively, experimental approaches have also been developed for unbiased genome-wide detection of off-target cleavages^66–72. Continued investigation is necessary to more completely understand the rules governing sgRNA efficiency and off-target mutagenesis (see Box 1 for further discussion).

CRISPResso indel analysis can supplement other common analyses, such as analyses of gene expression or protein-level changes. Although CRISPResso is useful for quantifying editing frequency to demonstrate that editing has successfully occurred and demonstrates the full indel spectrum (substitutions, insertions, and deletions), it is often helpful to perform other techniques to assess for changes beyond genomic DNA (gDNA) for further evaluation (i.e., gene expression or protein-level changes).

EQUIPMENT

• Corning untreated 245-mm² bioassay dishes (Fisher Scientific, cat. no. 431111)

• Ultracentrifuge tube, thin-wall, polypropylene, 38.5 ml, 25 × 89 mm (Beckman Coulter, cat. no. 326823)

• Falcon 50-ml conical centrifuge tubes (Fisher Scientific, cat. no. 14-432-22)

• Corning tissue culture (TC)-treated culture dishes, 15 cm, round (Fisher, cat. no. 08-772-24)

• Penicillin–streptomycin (Life Technologies, cat. no. 15140122)

• Solid glass beads (Fisher Scientific, cat. no. 11-312-10B)

• Electroporation system (e.g., Bio-Rad’s Gene Pulser MXcel, Gene Pulser Xcell, and MicroPulser electroporators)

• Corning Falcon bacteriological Petri dishes with lids (Fisher Scientific, cat. no. 08-757-100D)

• MiSeq or HiSeq sequencing system (Illumina) or equivalent sequencing platform

• Qubit fluorometer (Thermo Fisher Scientific)

• UV-visible spectrophotometer (NanoDrop)

• Gel visualization system (Alpha Innotech)

• Thermocycler (Bio-Rad)

• UV-light transilluminator (Fisher Scientific)

• UV-filter face mask (Fisher Scientific) ! CAUTION Wear gloves, lab coat, and face shield to avoid harm to eyes and skin caused by UV light.

• Microcentrifuge tubes, polypropylene (VWR, cat. no. 87003-294)

• 8-Strip PCR tubes, 0.2 ml (Fisher Scientific, cat. no. 14-222-250)

• Sterile culture tube with attached dual-position cap (VWR, cat. no. 89497-812)

• Ultracentrifuge (Beckman Coulter)

• SW32Ti rotor with compatible swinging buckets (Beckman Coulter, cat. no. 369694)

• Falcon cell scraper with 40-cm handle and 3.0-cm blade (Corning, cat. no. 353087)

COMPUTING EQUIPMENT

• Computer with at least 8 GB of RAM, 1 TB of disk space, and a recent central processing unit (CPU) with two or more cores (command-line and Docker version only). Windows (Docker only), OSX, and Linux platforms are supported.

• Anaconda Python 2.7 (https://www.continuum.io/downloads (CRISPResso is not compatible with Python 3))

• Docker (https://www.docker.com/community-edition#/download

• Sequence analysis software, e.g., free software such as ApE (http://biologylabs.utah.edu/jorgensen/wayned/ape/), SerialCloner (http://serialbasics.free.fr/Serial_Cloner.html), or commercial software such as Thermo Scientific Vector NTI (https://www.thermofisher.com/us/en/home/life-science/cloning/vector-nti-software.html), Lasergene DNAstar (https://www.dnastar.com/t-allproducts.aspx), CLC Main Workbench (https://www.qiagenbioinformatics.com/products/clc-main-workbench), or MacVector (http://macvector.com)

• CRISPResso command-line software (https://github.com/pinellolab/CRISPResso)

  ▲ CRITICAL Refer to the ‘Equipment Setup’ section for detailed installation instructions and troubleshooting advice for CRISPResso.

• CRISPOR command-line software: http://crispor.org/downloads

  ▲ CRITICAL Refer to the ‘Equipment Setup’ section for detailed installation instructions and troubleshooting advice for CRISPROR.

• FASTQC (http://www.bioinformatics.babraham.ac.uk/projects/fastqc/)

REAGENT SETUP

(A) sgRNA design for arrayed and pooled experiments using the CRISPOR webtool ● TIMING 1–4 h

1. Open the CRISPOR webtool (http://crispor.tefor.net/).

2. Input the target DNA sequence. The webtool requires a sequence input (< 2 kb) or genomic coordinates.

   ▲ CRITICAL STEP The input sequence must be a genomic sequence as opposed to a cDNA sequence, as the latter can include sequences that are not in the genome due to splicing. You can obtain genomic sequences using a website such as the UCSC Genome Browser (https://genome.ucsc.edu, click on ‘View - DNA’ after searching for a gene or genomic region) or Ensembl (http://www.ensembl.org, click on ‘Export data – Text’ after searching for a gene or genomic region). From the UCSC Genome Browser, the current sequence in view can be sent directly to CRISPOR via the menu entry ‘View – In external tools’.

   ? TROUBLESHOOTING

3. Select the relevant genome assembly (e.g., hg19, mm9) and PAM sequence for the nuclease (e.g., NGG for Streptococcus pyogenes Cas9).

   ▲ CRITICAL STEP It is important to pick the appropriate assembly to be consistent with the genomic coordinates of the DNA sequence provided in Step 1A(ii) and for accurate prediction of off-target sites.

4. Click on the ‘Submit’ button to run the CRISPOR analysis. Select the ‘Cloning/PCR primers’ link underneath each sgRNA sequence for automated primer design for future indel analysis by CRISPResso at on- and off-target loci. If you are conducting a saturating mutagenesis screen, the list of the oligonucleotide pool sequences for Gibson assembly, sequencing primers for validation, sequencing amplicons for CRISPResso, and the full list of sgRNA sequences can be downloaded by following the link ‘Saturating Mutagenesis Assistant’ at the top of the CRISPOR output page. The four files provided by the Saturating Mutagenesis Assistant can be downloaded by selecting from the drop-down menu adjacent to ‘output file’. The outputs are described in Table 2. Alternatively, the full list of sgRNA sequences for saturating mutagenesis can also be found by following the ‘Cloning/PCR primers’ link under each sgRNA sequence and then following the link provided under the ‘Saturating mutagenesis using all guides’ heading.

   Table 2

   Four output files from CRISPOR sgRNA design analysis.

   Output number	Filename	Description	File columns
   1	REGION_1_sat MutOligos.tsv	This file contains the sequences to order from a custom oligonucleotide pool supplier (MATERIALS)	GuideId: the identifier of the sgRNA sequence in the input sequence. It consists of the position of the PAM and the strand, e.g., ‘4rev’ targetSeq: the sgRNA sequence, including the PAM mitSpecScore: the MIT Guide Specificity score (0–100, higher score = lower off-target potential) off-target count: number of predicted off-target sites (by default at four mismatches) targetGenomeGeneLocus: gene symbol and sequence location name, e.g., ‘exon:PITX2’ Doench’16EffScore: the Doench 2016 guide efficiency score24 Moreno-MateosEffScore: the Moreno-Mateos 2015 (CRISPRscan) guide efficiency score52 OligoNucleotideAdapterHandle + PrimerFw: the forward oligonucleotide to order from a supplier OligoNucleotideAdapterHandle + PrimerRev: the reverse oligonucleotide to order from a supplier
   2	REGION_1_ontarget Primers.tsv	This file contains two primers for each sgRNA that can be ordered from an oligonucleotide supplier. The primers can be used to amplify the DNA fragment around each sgRNA	guideId: sgRNA identifier, see above forwPrimer: forward primer sequence forwPrimerTm: forward primer Tm revPrimer: reverse primer sequence revPrimerTm: reverse primer Tm ampliconSequence: the genomic sequence between forward and reverse primer
   3	REGION_1_ontarget Amplicons.tsv	After sequencing using the ontargetPrimers file, this file can be used as input	guideId: sgRNA identifier, see above ampliconSequence: the genomic sequence between forward and reverse primers, see above
   		For CRISPRessoPooled to determine the cleavage frequency of each sgRNA	guideSequence: the sgRNA sequence located within the amplicon
   4	REGION_1_tar getSeqs.tsv	This file contains a list of all sgRNA sequences (one per line). It can be used to quantify the relative abundance of these sgRNA in a sample of sequenced cells and is one of the input files for CRISPRessoCount	List of all sgRNA sequences (one per line)

   Open in a separate window

   ▲ CRITICAL STEP CRISPOR allows for PCR amplicon lengths of 50–600 bp. The melting temperature (Tm) is calculated for each primer pair.

(B) Gene-targeted library design generated from a gene list using the CRISPOR Batch gene-targeting assistant webtool ● TIMING 1–4 h

1. To generate a gene-targeted library using the CRISPOR Batch gene-targeting assistant, go to http://crispor.tefor.net/ and click on the ‘CRISPOR Batch’ link.

2. Select the source genome-wide library for the extraction of sgRNA sequences using the drop-down menu adjacent to ‘Lentiviral screen library’. References are provided for each library to assist with appropriate source library selection.

3. Select the desired ‘number of guides per gene’ (maximum of six sgRNAs/gene) using the drop-down menu. Note that some libraries have fewer than six sgRNAs per gene. If more sgRNAs are requested than are available for a given gene with the library, the maximum number of sgRNA sequences available within the library will be output instead.

4. Enter the ‘number of non-targeting control guides’ to include in the library (maximum = 1,000). See the ‘Experimental design’ section for further discussion of controls.

5. Select a barcode for library preparation as described in Step 16 (see ‘Synthesis of individual or multiple pooled sgRNA libraries’ within the ‘Experimental design’ section for further discussion of barcodes).

6. Copy and paste the gene list into the entry field. Genes should be entered as official gene symbols, with only one entry per line. Refer to the HUGO Gene Nomenclature Committee (HGNC) (https://www.genenames.org/) and Mouse Genome Informatics (MGI) (http://www.informatics.jax.org/) for official gene symbols for human and mouse, respectively.

   ▲ CRITICAL STEP This entry is case-insensitive because the libraries are already species-specific. If an entry is not found in the database of genome-wide sgRNAs, sgRNAs for the entered gene will display a warning and will not be included in the output.

   ? TROUBLESHOOTING

7. Click on ‘Submit’ for gene-targeted library generation and download the output library.

(C) sgRNA design using command-line CRISPOR ● TIMING 1–4 h

1. Type the following command into a terminal to download the latest version of the Docker container for this protocol:

   docker pull pinellolab/crispor_crispresso_nat_prot
   

2. Verify that the container was downloaded successfully by running the following command:

   docker run pinellolab/crispor_crispresso_nat_prot crispor.py
   

3. Obtain the CRISPOR assembly identifier of the genome (e.g., hg19 or mm9). If you are unsure, the full list of assemblies is available at http://crispor.tefor.net/genomes/genomeInfo.all.tab. This protocol uses ‘hg19’ in its examples; however, this should be switched to the relevant identifier for the assembly of interest as appropriate.

4. Execute the following commands to download the relevant pre-indexed genome in the folder created in the previous step. Here, we are assuming that the user will store the pre-indexed genomes in the folder /home/user/crispor_genomes.

   mkdir -p /home/user/crispor_genomes
   docker run -v /home/user/crispor_genomes:/crisporWebsite/genomes \
   pinellolab/crispor_crispresso_nat_prot downloadGenome hg19 \
   /crisporWebsite/genomes
   

5. Prepare the input genomic DNA (gDNA) sequences in FASTA format (see a description of FASTA format at https://www.ncbi.nlm.nih.gov/blast/fasta.shtml).

   ▲ CRITICAL STEP It is important to pick the appropriate genome assembly to be consistent with the genomic coordinates of the input DNA sequences. In particular, it is essential for accurate prediction of off-target sites.

6. (Optional) Generate a FASTA file with the exonic sequences. If you have a list of genes and you want to target their exons, go to the UCSC Genome Browser page http://genome.ucsc.edu/cgi-bin/hgTables.

7. Set the following options (refer to the description of each item in the ‘Table Browser’ to assist with setting each parameter). Of note, the ‘Table’ parameter will offer different options depending on the chosen ‘Track’. Step 1C(viii–xii) is based on the example entry from Step 1C(vii):

   UCSC Genome Browser table option	Example entry (adjust as appropriate)
   Clade	Mammal
   Genome	Human, assembly: hg19
   Group	Genes and gene predictions
   Track	Gencode V19
   Table	Basic (wgEncodeGencodeBasicV19)
   Output format	Sequence
   Output file	selected_genes_exons.fasta
   File type returned	‘plain text’

   Open in a separate window

8. Press the button ‘paste list’ and enter the list of transcript identifiers for each gene.

9. Press the button ‘get output’. A new page will open.

   ▲ CRITICAL STEP The ‘Send output to’ Galaxy, GREAT, or GenomeSpace checkboxes are off by default; however, they will remain checked if you have ever checked any of them in the past. Make sure that none of the checkboxes are selected.

10. Select the option ‘genomic’ and press ‘submit’. A new page will open.

11. Select ‘CDS: Exons’ and ‘One FASTA record per region (exon, intron, etc.)’.

12. Press the button: ‘get sequence’ and save the file as selected_genes_exons.fasta.

13. Run CRISPOR over your single- or multi-fasta input file by entering the following commands. Here, we assume that this file is stored in /home/user/crispor_data/crispor_input.fasta and contains a sequence with id > REGION_1.

    docker run \
    -v /home/user/crispor_data:/DATA \
    -v /home/user/crispor_genomes:/crisporWebsite/genomes \
    -w /DATA \pinellolab/crispor_crispresso_nat_prot \
    crispor.py hg19 crispor_input.fasta crispor_output.tsv --satMutDir=./
    

    The sgRNA sequences will be written to the file crispor_output.tsv and four files for each sequence ID in the FASTA file will be created as in the web version in the folder /home/user/crispor_data/ (see Table 2 for output files).

14. To select a subset of sgRNAs from the file REGION_1_satMutOligos.tsv as generated in Step 1A(iv) or 1C(xiii) (Table 2), select the lines corresponding to sgRNAs with the desired scores/attributes (Box 1) and save them to a new file called REGION_1_satMutOligos_filtered.tsv. Filter the sgRNAs by running the following commands:

    docker run -v /home/user/crispor_data/:/DATA -w /DATA pinellolab/
    crispor_crispresso_nat_prot bash -c "join -t $'\t' −1 1 −2 1
    REGION_1_satMutOligos_filtered.tsv REGION_1_ontargetAmplicons.tsv -o
    2.1,2.2,2.3 > CRISPRessoPooled_amplicons.tsv"
    docker run -v /home/user/crispor_data/:/DATA -w /DATA pinellolab/
    crispor_crispresso_nat_prot bash -c "join -t $'\t' −1 1 −2 1
    REGION_1_satMutOligos_filtered.tsv REGION_1_ontargetAmplicons.tsv -o 2.3
    | sed 1d > CRISPRessoCounts_sgRNA.tsv"
    docker run -v /home/user/crispor_data/:/DATA -w /DATA pinellolab/
    crispor_crispresso_nat_prot bash -c "join -t $'\t' −1 1 −2 1
    REGION_1_satMutOligos_filtered.tsv REGION_1_ontargetPrimers.tsv -o
    2.1,2.2,2.3,2.4,2.5,2.6,2.7 > REGION_1_ontargetPrimers_filtered.tsv"
    

    This will create a filtered version of the files described in Table 2 to use for designing amplicons and for performing CRISPResso analysis.

    ▲ CRITICAL STEP If the input file to CRISPOR was a multi-fasta file (Step 1C(xiii)), repeat this step for the other regions as appropriate.

Synthesis and cloning of individual sgRNA into a lentiviral vector ● TIMING 3 d to obtain sequence-confirmed, cloned sgRNA lentiviral plasmid

• 2|

  Individually resuspend lyophilized primer oligonucleotides at 100 µM in nuclease-free water.

  ■ PAUSE POINT Resuspended oligonucleotides can be stored at −20 °C for years.

• 3|

  Set up the phosphorylation reaction:

  Components	Amount (µl)	Final concentration
  Oligonucleotide 1 (at 100 µM)	1	–
  Oligonucleotide 2 (at 100 µM)	1
  10× Ligation buffer	1	1×
  T4 polynucleotide kinase (10 units/µl)	1	10 U
  Nuclease-free water	To 10

  Open in a separate window

• 4|

  Perform phosphorylation and annealing of the oligonucleotides in a thermocycler:

  Step	Temperature (°C)	Time	Comment
  1	37	30 min
  2	95	5 min
  3	95–25	30 s	Ramp down 5 °C/min

  Open in a separate window

  ■ PAUSE POINT Resuspended phosphorylated/annealed oligonucleotides can be stored at − 20 °C for years.

• 5|

  Prepare the digestion reaction by combining the following components in a 0.2-ml or 1.5-ml tube.

  Components	Amount	Final concentration
  lentiGuide-Puro plasmid	1 µg	–
  10× FastDigest buffer	2 µl	1×
  DTT (20 mM)	1 µl	1 mM
  Esp3l restriction enzyme	1 µl
  Nuclease-free water	To 20 µl

  Open in a separate window

• 6|

  Digest the sgRNA plasmid (lentiGuide-Puro) with Esp3l (an isoschizomer of BsmBI) restriction enzyme for 15 min at 37 °C.

• 7|

  Add 1 µl of TSAP to dephosphorylate the reaction for 15 min at 37 °C, followed by heat inactivation for 15 min at 74 °C.

• 8|

  Perform gel purification of the Esp3l-digested/dephosphorylated lentiGuide-Puro plasmid. See MATERIALS for gel purification kit and follow the manufacturer’s instructions. For further information on gel purification, refer to this previously published protocol⁸⁴.

• 9|

  Determine the concentration of the Esp3l-digested/dephosphorylated lentiGuide-Puro plasmid after gel purification using a UV-visible spectrophotometer (e.g., NanoDrop) or other equivalent method, following the manufacturer’s instructions.

• 10|

  Perform a ligation reaction for 5–15 min at room temperature using 30–70 ng of Esp3l-digested/dephosphorylated lentiGuide-Puro plasmid (generated in Steps 5–9) with the phosphorylated–annealed oligonucleotides produced in Steps 2–4 diluted 1:500 with nuclease-free water.

  Components	Amount	Final concentration
  2× Quick ligase reaction buffer	5 µl	1×
  Diluted oligonucleotides (from Step 4)	1 µl
  LentiGuide-Puro plasmid (from Steps 5 to 9)	30–70 ng	1×
  Quick ligase	1 µl	10 U
  Nuclease-free water	To 10 µl

  Open in a separate window

• 11|

  Heat-shock-transform the ligation reaction into stable competent E. coli by incubating samples at 42 °C for 45 s, followed by a 10-min incubation on ice. Add 250 µl of SOC medium and incubate at 37 °C with 250-r.p.m. shaking for 1 h. Use of recombination-deficient E. coli is important to minimize recombination events of repetitive elements in lentiviral plasmid (e.g., long terminal repeats). For more information about performing heat-shock transformation of competent E. coli, refer to previously published protocols^85,86.

• 12|

  Plate the transformation product from Step 11 on a 10-cm ampicillin-resistant (0.1 mg/ml ampicillin) agar plate. Incubate at 32 °C for 14–16 h. It is recommended that transformation products be incubated at 32 °C to reduce recombination between lentiviral long terminal repeats, but incubation at 37 °C can also be performed.

• 13|

  Picking three to five colonies is reasonable for individual incubation of each colony in 2 ml of ampicillin-resistant (0.1 mg/ml ampicillin) LB medium at 32 °C for 14–16 h with shaking at 250 r.p.m. It is recommended that transformation products be incubated at 32 °C to reduce recombination between lentiviral long terminal repeats, but incubation at 37 °C can also be performed.

  ? TROUBLESHOOTING

• 14|

  Perform mini-scale plasmid preparation of each 2-ml culture from Step 13 (see MATERIALS for a mini-scale plasmid preparation kit; follow the manufacturer’s instructions).

  ■ PAUSE POINT Successfully cloned sgRNA plasmids can be stored long term (years) at − 20 °C before creating lentivirus (Steps 49–71).

• 15|

  Sanger-sequence each colony-derived plasmid with the U6 sequencing primer to identify correctly cloned plasmids (see ‘Experimental design’ section for U6 primer sequence)⁶². Subsequent midi- or maxi-scale plasmid preparation may be required for lentivirus production (Steps 49–71).

Synthesis and cloning of pooled sgRNA libraries into a lentiviral vector ● TIMING 2–4 d to obtain cloned sgRNA lentiviral plasmid library

• 16|

  Use the sgRNAs from Step 1 to design full-length oligonucleotides (96–99 bp) flanked by barcodes and homologous sequence to the lentiGuide-Puro plasmid (Supplementary Table 2).

• 17|

  Order the designed DNA oligonucleotides on a programmable microarray (MATERIALS).

• 18|

  Obtain the resuspended oligonucleotides for use as the template for the PCR reaction in Step 19. The concentration will be determined by the manufacturer, so refer to the manufacturer for identifying the concentration.

  ■ PAUSE POINT Oligonucleotide pools can be stored at − 20 °C for short-term storage (months to years) and at − 80 °C for long-term storage (years).

• 19|

  Set up the barcode-specific PCR to amplify using the barcode-specific primers (Supplementary Table 3). This PCR will herein be referred to as ‘Library Synthesis PCR1’ (lsPCR1).

  ▲ CRITICAL STEP It is important to use a proofreading DNA polymerase to minimize introduction of PCR errors.

  Components	Amount (µl)	Final concentration
  Synthesized oligonucleotide template from Step 18	1	–
  5× Phusion HF buffer or 5× Q5 reaction buffer	10	1×
  Deoxynucleoside triphosphates (dNTPs; at 10 mM)	1	200 µM
  Barcode-specific forward primer (at 10 µM)	2.5	0.5 µM
  Barcode-specific reverse primer (at 10 µM)	2.5	0.5 µM
  Phusion Hot Start Flex DNA Polymerase or Q5 High-Fidelity DNA Polymerase	0.5	1.0 U/50-µl PCR
  Nuclease-free water	To 50

  Open in a separate window

  ? TROUBLESHOOTING

• 20|

  Perform multiple cycles (e.g., 10, 15, and 20 cycles) of lsPCR1 in a thermocycler as follows:

  Cycle number	Denature	Anneal	Extend
  1	98 °C, 30 s
  2-x (e.g., 10, 15, 20)	98 °C, 10 s	63 °C, 30 s	72 °C, 30 s
  x + 1			72 °C, 5 min

  Open in a separate window

  ▲ CRITICAL STEP The number of PCR cycles should be limited to reduce PCR bias. It is difficult to predict the required number of cycles for a given library. Therefore, it is important to perform multiple PCRs with different numbers of cycles to empirically determine the optimal number of cycles.

  ■ PAUSE POINT The PCR product can be stored for multiple days at 4 °C in the thermocycler at the end of the PCR program. Long-term storage (months to years) should be at − 20 °C.

• 21|

  Run a fraction (1–5 µl of the 50-µl PCR reaction from Step 20) of each lsPCR1 product with a different number of PCR cycles on a 2% (wt/vol) agarose gel. Analyze the gel to determine if the products occur at the expected size of ~100 bp (Supplementary Fig. 2a). If the bands run at the expected size, determine the minimal number of PCR cycles required to produce a visible band (Supplementary Fig. 2a). For example, 15 cycles would be used based on the gel presented in Supplementary Figure 2a; however, it may be possible to reduce the cycle number even further (i.e., another experiment with 5 and 10 PCR cycles). Use the remaining reaction volume for the cycle number chosen for Step 22 based on the gel from this step.

• 22|

  Using nuclease-free water, dilute the lsPCR1 reaction chosen in Step 21 1:10.

• 23|

  Set up ‘Library Synthesis PCR2’ (herein referred to as lsPCR2) with the universal lsPCR2 primers listed in Supplementary Table 4. lsPCR2 amplification removes the library-specific barcodes and adds a sequence homologous to the lentiGuide-Puro plasmid for Gibson assembly-based cloning of the library.

  ▲ CRITICAL STEP It is important to use a proofreading DNA polymerase to minimize introduction of PCR errors.

  Components	Amount (µl)	Final concentration
  Diluted lsPCR1 product from Step 22	1	–
  5× Phusion HF buffer or 5× Q5 reaction buffer	10	1×
  dNTPs (at 10 mM)	1	200 µM
  lsPCR2_forward primer (at 10 µM)	2.5	0.5 µM
  lsPCR2_reverse primer (at 10 µM)	2.5	0.5 µM
  Phusion Hot Start Flex DNA Polymerase or Q5 High-Fidelity DNA Polymerase	0.5	1.0 U/50-µl PCR
  Nuclease-free water	To 50

  Open in a separate window

• 24|

  Perform lsPCR2 using the same cycling conditions and multiple cycle numbers as Step 20.

  ■ PAUSE POINT The PCR product can be stored for multiple days at 4 °C in the thermocycler at the end of the PCR program. Long-term storage (months to years) should be at − 20 °C.

• 25|

  Run lsPCR2 products on a 2% (wt/vol) agarose gel to determine whether the products occur at the expected size of ~140 bp (Supplementary Fig. 2b). Choose the sample with the minimal number of PCR cycles that produces a visible band in order to minimize PCR bias (Supplementary Fig. 2b). For example, 15 cycles would be used based on the gel presented in Supplementary Figure 2b; however, it may be possible reduce cycle number even further (i.e., another experiment with 5 and 10 PCR cycles).

• 26|

  Gel-purify the ~140-bp band. See MATERIALS for gel purification kit and follow the manufacturer’s instructions. For further information on gel purification, refer to this previously published protocol⁸⁴.

  ▲ CRITICAL STEP If synthesizing multiple libraries simultaneously, leave empty lanes between different samples during gel electrophoresis to minimize the risk of sample contamination during gel purification.

• 27|

  Set up a Gibson assembly reaction using gel-purified Esp3l-digested/dephosphorylated lentiGuide-Puro plasmid from Steps 5 to 9:

  Components	Amount	Final concentration
  Gel-purified lsPCR2 from Step 26	10 ng
  Esp3l-digested/dephosphorylated lentiGuide-Puro plasmid from Steps 6 to 9	25 ng
  2× Gibson assembly master mix	10 µl	1×
  Nuclease-free water	To 20 µl

  Open in a separate window

• 28|

  Incubate the Gibson assembly reaction mix in a thermocycler at 50 °C for 60 min.

• 29|

  Thaw one 25-µl aliquot of electrocompetent bacterial cells on ice until they are completely thawed (~10–20 min).

• 30|

  Mix thawed electrocompetent bacterial cells by gently tapping on the side of the tube.

• 31|

  Add 25 µl of thawed electrocompetent bacterial cells to a prechilled 1.5-ml microcentrifuge tube on ice.

• 32|

  Add 1 µl of the Gibson assembly reaction mixture from Step 28 to 25 µl of thawed electrocompetent bacterial cells (from Step 31).

  ▲ CRITICAL STEP It is reasonable to aim for > 50× representation of the sgRNA library. Depending on the achieved transformation efficiency as calculated in Step 41 and the number of sgRNAs in the library, it may be necessary to set up more than 1 transformation reaction from Step 32. It is important to set up multiple reactions using 1 µl of Gibson assembly reaction mixture and 25 µl of thawed electrocompetent bacterial cells. Increasing the amount of Gibson assembly reaction mixture beyond 1 µl for a given electroporation can alter the chemistry of the electroporation and reduce transformation efficiency.

  ? TROUBLESHOOTING

• 33|

  Carefully pipette 25 µl of the electrocompetent bacterial cell/Gibson assembly reaction mixture from Step 32 into a chilled 0.1-cm-gap electroporation cuvette without introducing bubbles. Quickly flick the cuvette downward to deposit the cells across the bottom of the cuvette’s well.

  ▲ CRITICAL STEP Minimize introduction of air bubbles, as they can interfere with plasmid electroporation efficiency.

• 34|

  Using an electroporator, electroporate the sample at 25 µF, 200 Ω, and 1,500 V. The expected time constant during electroporation is ~4.7 (typical range of 4.2–4.8).

  ? TROUBLESHOOTING

• 35|

  Add 975 µl of recovery or SOC medium to the cuvette as soon as possible after electroporation (to a final volume of 1 ml in the cuvette). Pipette up and down enough times to resuspend the cells within the cuvette (probably a single pipetting up and down will be sufficient). Recovery medium is preferred here; however, SOC medium can be used in its place.

  ▲ CRITICAL STEP Electroporation is toxic to bacterial cells. Addition of recovery or SOC medium as quickly as possible enhances bacterial cell survival after electroporation.

• 36|

  Transfer the cell mixture from Step 35 (~1 ml) to a culture tube already containing 1 ml of SOC medium.

• 37|

  Shake each culture tube (containing a total of 2 ml of recovery/SOC medium) at 250 r.p.m. for 1 h at 37 °C.

  ▲ CRITICAL STEP If multiple transformations were set up in Step 32 using the same Gibson assembly reaction mixture, the 2-ml culture mixes generated in Step 37 can be pooled/mixed together so that one set of dilution plates is sufficient for calculation of transformation efficiency (Step 38).

• 38|

  Remove 5 µl from the 2-ml mixture (or 5 µl from a larger volume, if samples were pooled/mixed as described in Step 37) from Step 37 and add it to 1 ml of SOC medium. Mix well and plate 20 µl of the mixture (20,000× dilution) onto a prewarmed 10-cm ampicillin-resistant (0.1 mg/ml ampicillin) agar plate and then plate the remaining 200 µl (2,000× dilution) onto a separate 10-cm ampicillin-resistant (0.1 mg/ml ampicillin) agar plate. These two dilution plates can be used to estimate transformation efficiency, which will help ensure full library representation of the sgRNA plasmid library. It is reasonable to aim for > 50× representation of the sgRNA library.

  ▲ CRITICAL STEP Adjust the dilutions as necessary to obtain plates with a colony density that allows for accurate counting of the number of colonies on the 10-cm plates.

• 39|

  Plate the remaining 2-ml transformation mixture (or 2 ml from a larger volume, if samples were pooled/mixed) from Step 37 onto a prewarmed square 24.5-cm ampicillin-resistant (0.1 mg/ml ampicillin) agar plate using beads to ensure even spreading. Spread the liquid culture until it is largely absorbed into the agar and will not drip when inverted.

  ▲ CRITICAL STEP If multiple samples were mixed/pooled together in Step 37, it is still important to only plate 2 ml onto each square 24.5-cm ampicillin-resistant agar plate, as volumes > 2 ml may not be fully absorbed into the agar.

• 40|

  Invert all three (or more) plates (2,000× dilution plate, 20,000× dilution plate, and square 24.5-cm nondilution plate(s)) from Steps 38 and 39, and grow for 14–16 h at 32 °C. It is recommended to incubate transformations at 32 °C to reduce recombination between lentiviral long terminal repeats, but incubation at 37 °C can also be performed.

• 41|

  Count the number of colonies on the two dilution plates from Step 38. Multiply this number of colonies by the dilution factor (2,000× or 20,000×) and by the increased area of the square 24.5-cm plate (~7.6-fold increase in area) for estimation of the total number of colonies on the square 24.5-cm plate. It is reasonable to aim for > 50× representation of the sgRNA library.

  ? TROUBLESHOOTING

• 42|

  Select ~10–20 colonies from the dilution plates (from Step 41) for screening by mini-scale plasmid preparation and subsequent Sanger sequencing (as in Steps 14 and 15) to determine whether the sgRNA library has been PCR-amplified. Given that the oligonucleotides for multiple libraries can be synthesized on the same programmable microarray (Steps 16 and 17), it is important to confirm that the correct library intended for amplification has been obtained. This is also useful for preliminary evaluation of library representation. The expectation is to identify unique sgRNA sequences among all sequenced mini-scale plasmid preparations, given that the probability of obtaining the same sgRNA from a full library of sgRNAs should be low.

  ▲ CRITICAL STEP This step is intended to offer qualitative re-assurance that library synthesis is proceeding in the expected manner by determining that the correct library has been PCR-amplified and offering an initial assessment of library representation. Ultimately, the deep-sequencing step (Step 48) provides a more comprehensive assessment of library composition.

  ? TROUBLESHOOTING

• 43|

  Pipette 10 ml of LB medium onto each square 24.5-cm plate from Steps 39 and 40 and scrape the colonies off with a cell scraper. The LB medium aids in removal of the colonies from the agar.

• 44|

  Pipette the LB medium/scraped bacterial colony mixture into a preweighed 50-ml tube and repeat the procedure a second time on the same plate with an additional 5–10 ml of LB to maximize removal of bacterial colonies.

• 45|

  Centrifuge the LB medium/scraped bacterial colony mixture at 400g for 5 min at room temperature to pellet the bacteria and then discard the supernatant.

• 46|

  Weigh the bacterial pellet in the tube (and subtract the preweighed tube to determine the weight of the bacterial pellet) to determine the proper number of columns for maxi-scale plasmid preparation of the library. Each column can support ~0.45 g of bacterial pellet.

  ? TROUBLESHOOTING

• 47|

  Perform a sufficient number of maxi preps, following the manufacturer’s instructions and combine the eluted library plasmid DNA in a 1.5-ml tube.

• 48|

  (Optional) To confirm successful library representation, it is recommended to deep-sequence the batch-cloned plasmid library produced in Steps 16–47 by following the deep-sequencing procedure found in Steps 73–82 and using library plasmid DNA as the template DNA for laPCR1 in Step 75.

Lentivirus production from individual sgRNA plasmid or pooled sgRNA plasmid library ● TIMING 4 d

! CAUTION Take all necessary precautions for the handling of lentivirus and ensure proper disposal of lentiviral waste. Lentivirus is capable of integrating into the genome of human cells. This caution should be exercised during Steps 49–71.

• 49|

  Passage and maintain HEK293 cells in a 15-cm round plate with 16 ml of HEK293 medium as previously described⁶².

• 50|

  Perform transfection when HEK293 cells reach ~80% confluency in a 15-cm round plate using polyethylenimine (PEI) as a transfection reagent. The plasmid/PEI ratio should be 1 µg of the total transfected plasmid to 3 µg of PEI. Total plasmid consists of the sum of micrograms of VSV-G + micrograms of psPAX2 + library plasmid.

• 51|

  Mix PEI, VSV-G, psPAX2, and library plasmid in 1 ml of filtered DMEM without supplements in a sterile microcentrifuge tube:

  Reagent	Amount (µg)
  VSV-G plasmid	8.75
  psPAX2 plasmid	16.25
  Individual sgRNA plasmid or sgRNA library plasmid from Step 15 or 47	25
  Branched PEI (at 10 µg/µl)	150

  Open in a separate window

  ? TROUBLESHOOTING

• 52|

  Invert the microcentrifuge tube several times to mix. Allow the tube to incubate at room temperature for 20–30 min.

  ▲ CRITICAL STEP The DMEM/plasmid DNA/PEI mixture should change from translucent to opaque during this incubation period. If the color change does not occur, it is likely that one (or more) of the components is missing, and Steps 51 and 52 should be repeated.

• 53|

  Add the full volume (~1 ml) of DMEM/plasmid DNA/PEI mixture dropwise to the ~80% confluent HEK293 cells from Steps 49 and 50 and incubate the mixture at 37 °C for 16–24 h.

• 54|

  Replace the medium with 16 ml of fresh HEK293 medium 16–24 h after transfection in Step 53.

• 55|

  Lentiviral supernatant harvest no. 1.24 h after replacing the medium in Step 54, collect the 16 ml of medium that now contains lentiviral particles (herein referred to as ‘viral supernatant’) in a 50-ml tube. Replace the medium with 16 ml of fresh HEK293 medium. Store the viral supernatant at 4 °C until Step 56 is performed. The viral supernatant can be stored at 4 °C for up to 7 d before ultracentrifugation; however, storage at 4 °C may result in reduction of viral titer.

• 56|

  Lentiviral supernatant harvest no. 2.24 h after lentiviral supernatant harvest no. 1, collect the 16 ml of viral supernatant and put it into the same 50-ml tube from Step 55 (should now contain ~32 ml of viral supernatant). Appropriately discard the plate of HEK293 cells.

• 57|

  Centrifuge the collected viral supernatant from Step 56 at 500–700g for 5 min at 4 °C to pellet the HEK293 cells and other debris.

• 58|

  Filter the viral supernatant through a 0.45-µm 50-ml filter.

  ? TROUBLESHOOTING

• 59|

  Physical or functional titering can be performed. Physical titering involves detection of viral DNA (e.g., qPCR-based amplification of the viral genome, see MATERIALS) or viral proteins (e.g., ELISA for p24, see MATERIALS). Functional titering involves determination of the transduction rate using cells. Either physical or functional titering can be performed based on experimental needs. Refer to Kutner et al. for various titering protocols⁸⁷.

• 60|

  If sufficient viral titer is obtained with the viral supernatant from Step 59, the filtered viral supernatant can be stored at − 80 °C. It is recommended to freeze aliquots with volumes appropriate for future experimental needs to minimize freeze–thaw cycles, as they result in a reduction of viral titer. Aliquots also help to ensure consistent titer for each use of the same batch of virus. If insufficient viral titer is achieved based on the desired multiplicity of infection, it can be improved via ultracentrifugation (Steps 61–71).

  ■ PAUSE POINT The filtered viral supernatant can be stored at 4 °C for up to 7 d before ultracentrifugation; however, storage at 4 °C may result in reduction of viral titer. If ultracentrifugation is required (Steps 61–71), it is recommended to perform ultracentrifugation as soon as possible (minimize storage time at 4 °C). Lentiviruses can be stored at −80 °C for months. Longer storage will result in a decrease of viral titer.

(Optional) Ultracentrifugation of viral supernatant ● TIMING 2.5 h

• 61|

  Transfer the filtered viral supernatant to ultracentrifugation tubes.

• 62|

  Add 4–6 ml of 20% (wt/vol) sucrose solution to the bottom of the ultracentrifugation tube to create a sucrose layer (also known as a ‘sucrose cushion’) below the viral supernatant. This sucrose layer helps to remove any remaining debris from the viral supernatant, as only the high-density viral particles can pass through the sucrose layer to pellet, whereas the low-density debris remains in the supernatant.

  ? TROUBLESHOOTING

• 63|

  Add sterile PBS to the top of the ultracentrifuge tube until the final liquid volume within the ultracentrifuge tube is ~2–3 mm from the top.

  ▲ CRITICAL STEP The volume must be within 2–3 mm of the top of the tube to prevent the tube from collapsing because of the force of ultracentrifugation.

• 64|

  Place ultracentrifuge tubes into ultracentrifuge buckets.

• 65|

  Weigh all ultracentrifuge buckets containing ultracentrifuge tubes before ultracentrifugation to ensure appropriate weight-based balancing. Use sterile PBS to correct weight discrepancies.

• 66|

  Centrifuge at 100,000g for 2 h at 4 °C in an ultracentrifuge.

• 67|

  Remove ultracentrifuge tubes from ultracentrifuge buckets. The sucrose gradient should still be intact.

  ? TROUBLESHOOTING

• 68|

  Discard the supernatant by inverting the ultracentrifugation tube. Allow the inverted ultracentrifuge tube to stand on sterile paper towels to dry for 1–3 min.

  ? TROUBLESHOOTING

• 69|

  Revert the ultracentrifugation tube to an upright orientation and add 100 µl of sterile DMEM supplemented with 1% (vol/vol) FBS.

• 70|

  Incubate for 3–4 h (or overnight) on a rotator/rocker at ~10–15 rotations per minute at 4 °C to resuspend the lentiviral pellet.

  ■ PAUSE POINT Resuspension at 4 °C can be left overnight on a rotator/rocker.

• 71|

  Store concentrated lentivirus in a microcentrifuge tube at − 80 °C. Owing to loss of lentiviral titer with repeated freeze–thaw cycles, consider aliquoting at this point with volumes appropriate for experimental needs (i.e., aliquot size with enough lentivirus for one experiment). Aliquots frozen at the same time can be assumed to have the same titer, which can be advantageous for multiple experiments using the same lentivirus.

  ■ PAUSE POINT Lentiviruses can be stored at − 80 °C for months. Longer storage will result in a decrease of viral titer.

Execution of arrayed or pooled screen experiments ● TIMING 1–2 weeks

• 72|

  Perform arrayed or pooled screen experiments. Experiments/screens using lentivirus should be performed as appropriate for the experimental design/objectives and the cells being used for study. Refer to Boxes 3 and 4 for discussions about and considerations for performing an arrayed or pooled screen experiment.

  ■ PAUSE POINT After experiments/screens have been completed, cell pellets can be stored at − 20 or − 80 °C for weeks to months before processing for deep-sequencing analysis of the samples (Steps 73–82).

Deep sequencing of arrayed or pooled screen experiments ● TIMING 1 d

• 73|

  Extract gDNA from fresh cell pellets after completion of genome-editing screens/experiments or from frozen pellets from previous experiments. See MATERIALS for genomic DNA extraction kit and follow the manufacturer’s instructions.

  ▲ CRITICAL STEP For arrayed experiments intended for indel analysis by CRISPResso (Step 83), it is recommended to include a non-edited sample from the same cell type. This sample will be a useful negative control when performing CRISPResso indel analysis/enumeration. Specifically, a non-edited sample is useful in regard to optimization of the CRISPResso analysis parameters summarized in Box 2. Furthermore, a non-edited sample can be useful in identifying single-nucleotide polymorphisms (SNPs) or other variants present in the cells under study that can confound CRISPResso analysis. If a variant/indel is identified in the non-edited sample, the CRISPResso settings can be altered to account for the variant/indel during CRISPResso indel enumeration (Box 2).

• 74|

  Determine the concentration of each gDNA sample using a UV-visible spectrophotometer (e.g., NanoDrop) or other equivalent method using the manufacturer’s instructions.

  ▲ CRITICAL STEP The amount of gDNA can vary based on experimental needs for an arrayed experiment and based on the number of sgRNAs in the library for a pooled experiment. On average, a genome from a single cell is ~6.6 pg⁸⁸. Use adequate gDNA to represent the desired number of cells. For an arrayed experiment for indel analysis of a single locus using CRISPResso, > 1,000 cells at a given locus is a reasonable number. For a pooled sgRNA library experiment, 100–1,000× coverage of the sgRNAs in the library is reasonable. For example, 6.6 µg of gDNA is estimated to represent 1 million cells. For a library with 1,000 sgRNAs, 6.6 µg of gDNA would provide 1,000× coverage.

• 75|

  Set up the ‘Library Analysis PCR1’ reaction (herein referred to as laPCR1). Use lentiGuide-Puro-specific primers (Supplementary Table 5) to enumerate the sgRNAs present for enrichment/dropout analysis or locus-specific primers to quantitate indels (Supplementary Table 6). lentiGuide-Puro-specific primers are suggested in Supplementary Table 5; however, other primer pairs to PCR-amplify from the lentiGuide-Puro construct are possible.

  ▲ CRITICAL STEP It is important to use a proofreading DNA polymerase to minimize introduction of PCR errors.

  ▲ CRITICAL STEP Optimization of DMSO concentration may be required (typically 1–10% (vol/vol)). 8% (vol/vol) is used in the reaction below.

  ▲ CRITICAL STEP It is important to sequence the cloned plasmid library as generated in Steps 16–47. Sequencing of the plasmid library allows for confirmation of successful cloning/representation of the library and can be used as an initial time point for a dropout (‘depletion’) screen (see Box 4 for details of this type of screen). Given that plasmids contain many fewer base pairs than a full genome, ≥50 pg of plasmid DNA will be sufficient to represent a pooled plasmid library.

  ? TROUBLESHOOTING

  Components	Amount (µl)	Final concentration
  gDNA or plasmid DNA	X	–
  5× Herculase II reaction buffer	10	1×
  dNTPs (at 100 mM)	1	2 mM
  lentiGuide_forward or Locus_forward primer (at 10 µM)	2.5	0.5 µM
  lentiGuide_forward or Locus_reverse primer (at 10 µM)	2.5	0.5 µM
  DMSO	4	8%
  Herculase II Fusion DNA Polymerase	0.5
  Nuclease-free water	To 50

  Open in a separate window

• 76|

  Perform multiple cycles (e.g., 10, 15, and 20 cycles) of laPCR1 in a thermocycler as follows. Gradient PCR may be required to determine the optimal annealing temperature (Supplementary Fig. 2c).

  Cycle number	Denature	Anneal	Extend
  1	95 °C, 30 s
  2-x (e.g., 10, 15, and 20)	95 °C, 10 s	60 °C, 30 s	72 °C, 30 s
  x + 1			72 °C, 5 min

  Open in a separate window

  ? TROUBLESHOOTING

  ■ PAUSE POINT The PCR product can be stored at 4 °C in a thermocycler at the end of the PCR program for multiple days. Longer-term storage (months to years) should be at − 20 °C.

• 77|

  Run a fraction (1–5 µl of the 50-µl PCR reaction from Steps 75 and 76) of the laPCR1 product with a different number of PCR cycles on a 2% (wt/vol) agarose gel.

  ▲ CRITICAL STEP The number of PCR cycles should be limited to reduce PCR bias. It is difficult to predict the required number of cycles for a given library. Therefore, it is important to perform multiple PCRs with different cycle numbers to empirically determine the optimal number of cycles.

  ? TROUBLESHOOTING

• 78|

  Set up ‘Library Analysis PCR2’ (herein referred to as laPCR2) using the barcode-specific primers (Supplementary Tables 7 and 8). Each sample will need a unique Illumina Nextera index to allow demultiplexing of multiple samples sequenced together. Set up two PCRs for each sample.

  ▲ CRITICAL STEP Two separate 10-µl reactions are performed, as opposed to a single 20-µl reaction, to minimize PCR bias.

  Components	Amount (µl)	Final concentration
  laPCR1 product (diluted 1:5 in nuclease-free water) from Steps 75–77	1	-
  5× Herculase II reaction buffer	2	1×
  dNTPs (at 100 mM)	0.1	1 mM
  Forward primer (at 2 µM)	1	0.2 µM
  Reverse primer (at 2 µM)	1	0.2 µM
  Herculase II Fusion DNA Polymerase	0.1
  Nuclease-free water	To 10

  Open in a separate window

• 79|

  Perform laPCR2 using the same cycling conditions and multiple cycle numbers as in Step 76.

  ▲ CRITICAL STEP The number of PCR cycles should be limited to reduce PCR bias. It is difficult to predict the required number of cycles for a given library. Therefore, it is important to perform multiple PCRs with different cycle numbers to empirically determine the optimal number of cycles.

  ■ PAUSE POINT The PCR product can be stored at 4 °C in a thermocycler at the end of the PCR program for multiple days. Longer-term storage (months to years) should be at − 20 °C.

• 80|

  Combine both 10-µl reactions from the same sample for a total volume of 20 µl. Run laPCR2 products on a 2% (wt/vol) agarose gel to determine whether the products occur at the expected size (expected size varies based on primers used in laPCR1). Choose the sample with the minimal number of PCR cycles that produces a visible band in order to minimize PCR bias (Supplementary Fig. 2d). For example, 20 cycles would be used based on the gel presented in Supplementary Fig. 2d; however, it may be possible to reduce the cycle number even further (i.e., another experiment with 5, 10, and 15 cycles). Then gel-purify the relevant band of the expected size.

  ▲ CRITICAL STEP Leave empty lanes between different samples to minimize the risk of sample contamination during gel purification.

• 81|

  Quantitate DNA concentration by Qubit or another equivalent method, following the manufacturer’s instructions. Refer to the manufacturer’s instructions for the sequencing platform to be used (Step 82) for the minimum amount of DNA required. If an inadequate amount of DNA is obtained, repeat Steps 75–81.

• 82|

  Sequence the libraries on a MiSeq, HiSeq, or equivalent sequencing platform (see ‘Experimental design’ for details).

  ▲ CRITICAL STEP It can be useful to discuss experimental goals with the sequencing facility or individual performing the sequencing prior to sequencing samples to help ensure that the desired number of reads are obtained to adequately represent the sequenced locus/loci or sgRNA library.

CRISPResso analysis of deep-sequencing data ● TIMING 10–120 min

• 83|

  The CRISPResso utility can be used for analysis of deep-sequencing data of a single locus/amplicon on a local machine or proprietary server, using a command-line version or a webtool freely available at http://crispresso.rocks/. Installation instructions are provided in the ‘Equipment Setup’ section for command-line and Docker versions. No installation is required for the webtool. The following step will present detailed instructions for workflows using the webtool (option A), command-line (option B), or Docker (option C) versions:

(A) Analysis of deep-sequencing data using the CRISPR esso webtool ● TIMING 10–120 min

▲ CRITICAL An illustrative video of the entire process is provided here: http://www.youtube.com/embed/dXblIIiAe00?autoplay=1.

1. Open a web browser and go to the CRISPResso page http://crispresso.rocks/.

2. Select the option for single-end (one FASTQ file) or paired-end reads (two FASTQ files) based on the type of deep sequencing performed. Analysis of paired-end reads requires overlapping sequence, as the paired reads will be merged. Upload the relevant FASTQ file(s) (as .fastq or .fastq.gz) for analysis. An example data set to test the tool with previously generated FASTQ files is provided at http://crispresso.rocks/help (see the ‘Try it now!’ section).

   ? TROUBLESHOOTING

3. Input the amplicon sequence (required), sgRNA sequence (optional), and coding sequence(s) (optional input for when targeting exonic sequence). The reads uploaded in the previous step will be aligned to the provided amplicon sequence. If an sgRNA sequence is provided, its position will be indicated in all output analyses. If a coding sequence is provided, it will allow CRISPResso to perform frameshift analysis. The provided exonic sequence must be a subsequence of the amplicon sequence and not the sequence of the entire exon.

   ? TROUBLESHOOTING

4. Adjust the parameters for window size (bp around each side of cleavage site) to quantify NHEJ edits (if sgRNA sequence is provided), minimum average read quality (Phred33 scale), minimum single-bp quality (Phred33 scale), exclude bp from the left side of the amplicon sequence for the quantification of the mutations, exclude bp from the right side of the amplicon sequence for the quantification of the mutations, and trimming adapter (see Box 2 for a discussion of these parameters).

5. Submit samples for analysis and download analysis reports.

(B) Analysis of deep-sequencing data using command-line CRISPR esso ● TIMING 10–120 min

1. Gather the required information and files (see Step 1A(iv) or 1C(xiii); the CRISPOR utility will output all the required information to run CRISPResso).

   Input	Example entry (adjust as appropriate) or description
   FASTQ files for analysis (.fastq or .fastq.gz)	reads1.fastq.gz reads2.fastq.gz
   Reference amplicon sequence	AATGTCCCCCAATGGGAAGTTCATCTGGCACTGCCCACAGGTGA
   	GGAGGTCATGATCCCCTTCTGGAGCTCCCAACGGGCCGTGGTCT
   	GGTTCATCATCTGTAAGAATGGCTTCAAGAGGCTCGGCTGTGGTT
   sgRNA(s) used	TGAACCAGACCACGGCCCGT
   Expected repaired amplicon sequence	Only provide this if it is necessary to quantify HDR efficiency
   Desired single-base-pair or average read quality in Phred33 score	We suggest using 20 for single-base-pair quality and 30 for the average read quality
   Exonic sequence (for frameshift analysis)	Only provide this if the reference amplicon sequence contains exonic sequence
   Experiment name	BCL11A_exon2

   Open in a separate window

2. Open a terminal and run the CRISPResso utility by running the following command. Here, we are assuming that the FASTQ files reads1.fastq.gz and reads2.fastq.gz are stored in the folder /home/user/amplicons_data/.

   CRISPResso \
   -r1 /home/user/amplicons_data/reads1.fastq.gz \
   -r2 /home/user/amplicons_data/reads2.fastq.gz \
   -a AATGTCCCCCAATGGGAAGTTCATCTGGCACTGCCCACAGGTGAGG\
   AGGTCATGATCCCCTTCTGGAGCTCCCAACGGGCCGTGGTCTGGTTCAT\
   CATCTGTAAGAATGGCTTCAAGAGGCTCGGCTGTGGTT \
   -g TGAACCAGACCACGGCCCGT \
   -s 20 \
   -q 30 \
   -n BCL11A_exon2
   

   ▲ CRITICAL STEP Be sure to check whether or not the sequencing files were trimmed for adapters (refer to Box 2 for further discussion of the trimming process and options).

   ▲ CRITICAL STEP The command is based on the example data from Step 83B(i) and must be amended before use. After the execution of the command, a new folder with all the results will be created; in this example, the folder will be in /home/user/amplicons_data/CRISPResso_on_BCL11A_exon2. A summary of the different events discovered in the sequencing data is presented in the Alleles_frequency_table.txt file, and in several illustrative plots (.pdf files). Refer to the online documentation for the full description of the output: https://github.com/pinellolab/CRISPResso.

   ? TROUBLESHOOTING

(C) Analysis of deep-sequencing data using CRISPR esso with Docker ● TIMING 10–120 min

1. Gather the required information and files as in Step 83B(i).

2. Assuming that the FASTQ files reads1.fastq.gz and reads2.fastq.gz are stored in the folder /home/user/amplicons_data/, run the command

   docker run \
   -v /home/user/amplicons_data:/DATA \
   -w /DATA pinellolab/crispor_crispresso_nat_prot \
   CRISPResso \
   -r1 reads1.fastq.gz \
   -r2 reads2.fastq.gz \
   -a AATGTCCCCCAATGGGAAGTTCATCTGGCACTGCCCACAGGTGAGG\
   AGGTCATGATCCCCTTCTGGAGCTCCCAACGGGCCGTGGTCTGGTTCAT\
   CATCTGTAAGAATGGCTTCAAGAGGCTCGGCTGTGGTT \
   -g TGAACCAGACCACGGCCCGT \
   -s 20 \
   -q 30 \
   -n BCL11A_exon2
   

   After the execution of the command, a new folder with all the results will be created; in this example, the folder will be in /home/user/amplicons_data/CRISPResso_on_BCL11A_exon2. A summary of the different events discovered in the sequencing data is presented in the Alleles_frequency_table.txt file, and in several illustrative plots (.pdf files). Refer to the online documentation for the full description of the output: https://github.com/pinellolab/CRISPResso.

   ? TROUBLESHOOTING

Analysis of a pooled sgRNA experiment using CRISPRessoCount ● TIMING 20 min–4 h

▲ CRITICAL CRISPRessoCount is a utility for the enumeration of sgRNA (Fig. 2; Box 4). It is necessary to use the command-line or Docker version of CRISPResso to run CRISPRessoCount. CRISPRessoCount can enumerate sgRNAs from a user-generated list or can empirically identify all sgRNAs present in the FASTQ file.

• 84|

  Gather the required data and information (adjust the entries as appropriate for the experiment):

  Input	Example entry (adjust as appropriate) or description
  FASTQ files for sgRNA enumeration analysis (.fastq or .fastq.gz)	sample_screen.fastq.gz
  File containing only the sgRNA sequences to enumerate, one per line. This is an optional input. If not provided, all unique sgRNA sequences will be enumerated	library_NGG.txt (or the file created by CRISPOR, see Step 1A(iv) or 1C(xiii))
  The sgRNA scaffold sequence immediately downstream of the cloned sgRNA sequence	GTTTTAGAGCTAGAAATAGC
  sgRNA length	20
  Desired single-base-pair or average read quality in Phred33 score	We suggest 20 for single-base-pair quality and 30 for the average read quality
  Experiment name	SAMPLE_SCREEN

  Open in a separate window

• 85|

  Run the CRISPRessoCount analysis. The following step will present detailed instructions for running CRISPRessoCount from the command line (option A) or Docker (option B). Here, we are assuming that the FASTQ file and the optional sgRNA file are stored in the folder home/user/sgRNA _data/.

(A) Using the command-line version for CRISPR essoCount analysis ● TIMING 10–120 min

1. Run the CRISPRessoCount analysis with the following commands:

   CRISPRessoCount -r /home/user/sgRNA_data/sample_screen.fastq.gz \
   -s 20 -q 30 -f library_NGG.txt \
   -t GTTTTAGAGCTAGAAATAGC -l 20 --name SAMPLE_SCREEN
   

(B) Using the Docker container for CRISPR essoCount analysis ● TIMING 10–120 min

1. Run the CRISPRessoCount analysis with the following commands:

   docker run \
   -v /home/user/sgRNA_data:/DATA \
   -w /DATA \
   pinellolab/crispor_crispresso_nat_prot \
   CRISPRessoCount -r sample_screen.fastq.gz \
   -s 20 -q 30 -f library_NGG.txt \
   -t GTTTTAGAGCTAGAAATAGC -l 20 --name SAMPLE_SCREEN
   

After the execution of the command, a new folder with all the results will be created; in this example, the folder will be in /home/user/amplicons_data/CRISPRessoCount_on_SAMPLE_SCREEN. This folder contains two files, the execution log (CRISPRessoCount_RUNNING_LOG.txt) and a file (CRISPRessoCount_only_ref_guides_on_sample_screen.fastq.gz.txt) containing a table with the raw and normalized counts for each sgRNA in the input FASTQ file:

• 86|

  Calculate the enrichment and/or dropout (‘depletion’) between two conditions, and calculate the log2 ratio using the normalized sgRNA counts (this takes into account the total number of reads within each sequenced sample) using the ‘Reads_Per_Millions_(RPM)’ column for the two conditions. Refer to the ‘CRISPResso: Analysis of deep sequencing from arrayed or pooled sgRNA experiments’ section for further discussion of read normalization.

  ▲ CRITICAL STEP The log₂ transformation is not necessary, but is often used when displaying the data, as it will center the data at ~0 for an equal ratio; positive values will represent enrichment and negative values will represent depletion/dropout.

  ? TROUBLESHOOTING

CRISPR esso analysis using CRISPRessoPooled ● TIMING 20 min–4 h

▲ CRITICAL CRISPRessoPooled is a utility for analyzing and quantifying targeted sequencing CRISPR experiments involving sequencing with pooled amplicons. To use CRISPRessoPooled, it is necessary to use the command-line version of CRISPResso (see Step 83B). Although CRISPRessoPooled can be run in different modes (amplicon only, genome only, and mixed), in this protocol, we use the most reliable mode for quantification, called ‘mixed mode’. In this mode, the amplicon reads are aligned to both reference amplicons and the genome to discard ambiguous or spurious reads. For more details regarding the different running modes, consult the online help (help is located below the list of files at the following link): https://github.com/pinellolab/CRISPResso.

• 87|

  Gather the required data and information, including FASTQ files that contain reads for a pooled experiment with multiple sgRNAs, for example:

  Input	Example entry (adjust as appropriate) or description
  FASTQ files for analysis (.fastq or .fastq.gz)	Reads_pooled_1.fastq.gz
  	Reads_pooled_2.fastq.gz
  Reference amplicon sequences used in the pooled experiment
  sgRNA sequences used, one for each amplicon
  Expected repaired amplicon sequence	Only provide this if it is necessary to quantify HDR efficiency
  Desired single-base-pair or average read quality in Phred33 score	We suggest 20 for single-base-pair quality and 30 for the average read quality
  Exonic sequence (for frameshift analysis)	Only provide this if the reference amplicon sequence contains exonic sequence
  Amplicon names	All the names should be unique
  A global name for the pooled analysis	Pooled_amplicons

  Open in a separate window

  ▲ CRITICAL STEP Be sure to check whether or not the sequencing files were trimmed for adapters (refer to Box 2 for further discussion of the trimming options).

• 88|

  Create a folder for the reference genome and genomic annotations (this step is necessary only one time):

  mkdir -p /home/user/crispresso_genomes/hg19
  

• 89|

  Download the desired genome sequence data and precomputed index from https://support.illumina.com/sequencing/sequencing_software/igenome.html. In this example (Steps 90 and 91), we download the hg19 assembly of the human genome.

• 90|

  Move to the folder containing the genome:

  cd /home/user/crispresso_genomes/hg19
  wget ftp://igenome:[email protected]\
  /Homo_sapiens/UCSC/hg1938/Homo_sapiens_UCSC_hg19.tar.gz
  

• 91|

  Extract the data with the following command:

  tar -xvzf Homo_sapiens_UCSC_hg19.tar.gz \
  Homo_sapiens/UCSC/hg19/Sequence/Bowtie2Index \
  --strip-components 5
  

• 92|

  Create an amplicon description file or use the amplicon file created by CRISPOR in Step 1A(iv) or 1C(xiii). This file is a tab-delimited text file with up to five columns (the first two columns are required) (Supplementary Fig. 3a):

  Column no.	Description
  1	An identifier for the amplicon (must be unique)
  2	Amplicon sequence used in the design of the experiment
  3 (optional)	sgRNA sequence used for this amplicon, without the PAM sequence. If not available, enter ‘NA’
  4 (optional)	Expected amplicon sequence in the case of HDR. If more than one, separate by commas and not spaces. If not available, enter ‘NA’
  5 (optional)	Subsequence(s) of the amplicon corresponding to coding sequences. If more than one, separate by commas and not spaces. If not available, enter ‘NA’

  Open in a separate window

  Here, we are assuming that the file is saved in the folder /home/user/pooled_amplicons_data/ and is called AMPLICONS_FILE.txt

• 93|

  Open a terminal and run the CRISPRessoPooled utility from the command line (option A) or Docker (option B). Here, we are assuming that the FASTQ files Reads_pooled_1.fastq.gz and Reads_pooled_1.fastq.gz are stored in the folder /home/user/pooled_amplicons_data/

(A) Using the command-line version for CRISPR essoPooled analysis ● TIMING 10–120 min

1. Run CRISPRessoPooled by executing the following commands:

   CRISPRessoPooled \
   -r1 /home/user/pooled_amplicons_data/Reads_pooled_1.fastq.gz \
   -r2 /home/user/pooled_amplicons_data/Reads_pooled_2.fastq.gz \
   -f /home/user/amplicons_data/AMPLICONS_FILE.txt \
   -x /home/user/crispresso_genomes/hg19/genome \
   -s 20 -q 30 \
   --name Pooled_amplicons
   

   ? TROUBLESHOOTING

(B) Using the Docker container for CRISPR essoPooled analysis ● TIMING 10–120 min

1. Run CRISPRessoPooled by executing the following commands:

   docker run \
   -v /home/user/amplicons_data/:/DATA -w /DATA \
   -v /home/user/crispresso_genomes/:/GENOMES \
   pinellolab/crispor_crispresso_nat_prot \
   CRISPRessoPooled \
   -r1 Reads_pooled_1.fastq.gz \
   -r2 Reads_pooled_2.fastq.gz \
   -f AMPLICONS_FILE.txt \
   -x /GENOMES/hg19/genome \
   -s 20 -q 30 \
   --name Pooled_amplicons
   

   After the execution of the command in the previous step, a new folder with all the results will be created. In this example, the folder will be in /home/user/amplicons_data/CRISPRessoPooled_on_Pooled_amplicons. In this folder, the user can find the following files:

   Output	Description
   REPORT_READS_ALIGNED_TO_GENOME_AND_AMPLICONS.txt	This file contains the same information provided in the input description file, plus some additional columns: Amplicon_Specific_fastq.gz_filename: name of the file containing the raw reads recovered for the amplicon n_reads: number of reads recovered for the amplicon chr_id: chromosome of the amplicon in the reference genome bpstart: start coordinate of the amplicon in the reference genome bpend: end coordinate of the amplicon in the reference genome Reference_Sequence: sequence in the reference genome for the region mapped for the amplicon
   MAPPED_REGIONS (folder)	This folder contains all the .fastq.gz files for the discovered regions
   Set of folders with CRISPResso reports	CRISPResso analysis for the recovered amplicons with enough reads
   SAMPLES_QUANTIFICATION_SUMMARY.txt	This file contains a summary of the quantification and the alignment statistics for each region analyzed (read counts and percentages for the various classes: unmodified, NHEJ, and HDR)
   CRISPRessoPooled_RUNNING_LOG.txt	Execution log and messages for the external utilities called

   Open in a separate window

   ? TROUBLESHOOTING

CRISPResso analysis using CRISPRessoWGS ● TIMING 20 min–4 h

• 94|

  Gather the required data and information:

  Input	Example entry (adjust as appropriate) or description
  Genome-aligned WGS data in BAM format	sample_WGS.bam
  The sequence of the reference genome used for the alignment, in FASTA format (Steps 88–91)	Human hg19 genome assembly in FASTA format
  (Optional) gene annotation file (Step 1C(vi–xii))	gencode_v19.gz
  Coordinates of the regions to analyze	REGIONS_FILE.txt
  Expected repaired amplicon sequence	Only provide this if it is necessary to quantify HDR efficiency
  Desired single-base-pair or average read quality in Phred33 score	We suggest 20 for single-base-pair quality and 30 for the average read quality
  Exonic sequences (for frameshift analysis)	Only provide this if the reference amplicon sequence contains exonic sequence
  Region names	All the names should be unique
  A global name for the WGS analysis	WGS_regions

  Open in a separate window

• 95|

  Create a regions description file. This file is a tab-delimited text file with up to seven columns (four are required) and contains the coordinates of the regions to analyze and some additional information (Supplementary Fig. 3b):

  Column no.	Description
  1	Chromosome of the region in the reference genome
  2	Start coordinate of the region in the reference genome
  3	End coordinate of the region in the reference genome
  4	An identifier for the region (must be unique)
  5 (optional)	sgRNA sequence used for this genomic segment without the PAM sequence. If not available, enter ‘NA’
  6 (optional)	Expected genomic segment sequence in case of HDR. If more than one, separate by commas and not spaces. If not available, enter ‘NA’
  7 (optional)	Subsequence(s) of the genomic segment corresponding to coding sequences. If more than one, separate by commas and not spaces. If not available, enter ‘NA’

  Open in a separate window

  Here, we are assuming that the file is saved in the folder /home/user/wgs_data/ and it is called REGIONS_FILE.txt.

• 96|

  Open a terminal and run the CRISPRessoWGS utility. The following step will present detailed instructions for the command-line version (option A) and Docker (option B). Here, we are assuming that the BAM file is stored in the folder /home/user/wgs_data/ and was aligned using the human hg19 reference genome.

(A) Using the command-line version for CRISPR essoWGS analysis ● TIMING 10–120 min

1. Run CRISPRessoWGS by executing the following commands:

   CRISPRessoWGS\
   -b /home/user/wgs_data/sample_WGS.bam \
   -f /home/user/wgs_data/REGIONS_FILE.txt \
   -r /home/user / crispresso_genomes/hg19/hg19.fa \
   -s 20 -q 30 \
   --name WGS_regions
   

   ? TROUBLESHOOTING

(B) Using the Docker container version for CRISPR essoWGS analysis ● TIMING 10–120 min

1. Run CRISPRessoWGS by executing the following commands:

   docker run \
   -v /home/user / wgs_data/:/DATA -w /DATA \
   -v /home/user / crispresso_genomes/:/GENOMES \
   pinellolab/crispor_crispresso_nat_prot \
   CRISPRessoWGS \
   -b sample_WGS.bam \
   -f REGIONS_FILE.txt \
   -r /GENOMES/hg19.fa \
   -s 20 -q 30 \
   --name WGS_regions
   

   After the execution of the command, a new folder with all the results will be created; in this example, the folder will be in /home/user/wgs_data/CRISPRessoWGS_on_WGS_regions. In this folder, the user can find the following files:

   Output filename	Description
   REPORT_READS_ALIGNED_TO_SELECTED_REGIONS_WGS.txt	This file contains the same information provided in the input description file, plus some additional columns: sequence: sequence in the reference genome for the region specified n_reads: number of reads recovered for the region bam_file_with_reads_in_region: file containing only the subset of the reads that overlap, also partially, with the region. This file is indexed and can be easily loaded, for example, onto the Integrative Genomics Viewer (IGV; http://software.broadinstitute.org/software/igv/) for visualization of single reads or for the comparison of two conditions fastq.gz_file_trimmed_reads_in_region: file containing only the subset of reads fully covering the specified regions and trimmed to match the sequence in that region. These reads are used for the subsequent analysis with CRISPResso
   ANALYZED_REGIONS (folder)	This folder contains all the BAM and FASTQ files, one for each region analyzed
   A set of folders with the CRISPResso report on the regions provided in the input	Refer to description of the output of CRISPResso (http://software.broadinstitute.org/software/igv/)
   CRISPRessoWGS_RUNNING_LOG.txt	Execution log and messages for the external utilities called

   Open in a separate window

   ? TROUBLESHOOTING

Direct comparison of CRISPResso analyses using CRISPRessoCompare and CRISPRessoPooledWGSCSCompare ● TIMING 20 min–4 h

• 97|

  Gather the required data and information: two completed CRISPResso, CRISPRessoPooled or CRISPRessoWGS analyses:

  Input	Example entry (adjust as appropriate) or description
  Path to two completed CRISPResso, CRISPRessoPooled, or CRISPRessoWGS analyses	/home/user/results/CRISPResso_on_sample1
  	/home/user/results/CRISPResso_on_sample2
  	or
  	home/user/results/CRISPRessoPooled_on_sample1
  	home/user/results/CRISPRessoPooled_on_sample2
  	or
  	/home/user/results/CRISPRessoWGS_on_sample1
  	/home/user/results/CRISPRessoWGS_on_sample2
  A name to use for the report	comparison_sample1_sample2

  Open in a separate window

• 98|

  Open a terminal and run the CRISPRessoCompare or CRISPRessoPooledWGSCompare utility from the command line (option A) or Docker (option B). Here, we are assuming that output folders from those utilities were saved in /home/user/results/.

(A) Using the command-line version for CRISPRessoCompare or CRISPRessoPooledWGSCSCompare analysis ● TIMING 10–120 min

1. Execute the following command for CRISPRessoCompare:

   CRISPRessoCompare \
   /home/user/results/CRISPResso_on_sample1 \
   /home/user/results/CRISPResso_on_sample2 \
   -n comparison_sample1_sample2
   

   Or execute the following command for CRISPRessoPooledWGSCompare:

   CRISPRessoPooledWGSCompare \
   /home/user/results/CRISPRessoPooled_on_sample1 \
   /home/user/results/CRISPRessoPooled_on_sample2 \
   -n comparison_sample1_sample2
   

M.C.C. was supported by a National Institute of Diabetes and Digestive and Kidney Diseases (NIDDK) Award (F30DK103359). M.H. was funded by National Institutes of Health (NIH)/National Human Genome Research Institute (NHGRI) grant 5U41HG002371-15 and NIH/National Cancer Institute (NCI) grant 5U54HG007990-02 and by a grant from the California Institute of Regenerative Medicine, CIRM GC1R-06673C. D.E.B. was supported by the National Heart, Lung, and Blood Institute (NHLBI) (DP2OD022716, P01HL032262), the Burroughs Wellcome Fund, and a Doris Duke Charitable Foundation Innovations in Clinical Research Award. S.H.O. was supported by an award from the NHLBI (P01HL032262) and an award from the NIDDK (P30DK049216, Center of Excellence in Molecular Hematology). N.E.S. was supported by the NIH through the NHGRI (R00-HG008171). L.P. was supported by an NHGRI Career Development Award (R00HG008399) and the Defense Advanced Research Projects Agency (HR0011-17-2-0042).