Genome Editing AnaLysis of UNidirectional sequencing for GenOme rearrangements
CRISPRlungo enables the analysis of single-anchor amplicon sequencing data through quantifying complex genome editing outcomes, identifying novel cut points, and using a biologically-aware alignment method to precisely measure small insertions and deletions.
A list of CRISPRlungo parameters can be accessed by runnning python CRISPRlungo.py -h. Parameters can be specified via the command line or a settings file. Parameters provided in the settings file will override settings provided on the command line in case of conflicts.
If provided, the settings file should be passed as the first argument to CRISPRlungo. The settings file may contain comments (lines starting with a '#' character). Setting names should be tab-separated from setting values.
- Download the example input file cuts.fq
- Prepare a Bowtie2 index of your genome or download the test genome folder
- In the same directory, create a settings file called
settings.txtwith the contents:
fastq_r1 cuts.fq
genome {path to genome}/Bowtie2Index/genome.fa or genome.chr11_5225364_5225863/genome.fa
guide_sequences CTTAGGGAACAAAGGAACCT
n_processes 20
primer_seq TTGCAATGAAAATAAATGTTT
Run CRISPRlungo with the command python CRISPRlungo.py settings.txt. View the example output by viewing the settings.txt.CRISPRlungo.html file.
settings_file Tab-separated settings file (default: None)
-h, --help show this help message and exit
-v, --version show program's version number and exit
--debug Print debug output (default: False)
--root ROOT, --name ROOT
Output directory file root (default: None)
--keep_intermediate If true, intermediate files are not deleted (default:
False)
--write_discarded_read_info
If true, a file with information for discarded reads
is produced (default: False)
--suppress_plots If true, no plotting will be performed (default:
False)
--guide_sequences [GUIDE_SEQUENCES ...]
Spacer sequences of guides (multiple guide sequences
are separated by spaces). Spacer sequences must be
provided without the PAM sequence, but oriented so the
PAM would immediately follow the provided spacer
sequence (default: [])
--cut_classification_annotations [CUT_CLASSIFICATION_ANNOTATIONS ...]
User-customizable annotations for cut products in the
form: chr1:234:left:Custom_label (multiple annotations
are separated by spaces) (default: [])
--cleavage_offset CLEAVAGE_OFFSET
Position where cleavage occurs, for in-silico off-
target search (relative to end of spacer seq -- for
Cas9 this is -3) (default: -3)
--genome GENOME Genome sequence file for alignment. This should point
to a file ending in ".fa", and the accompanying index
file (".fai") should exist. (default: None)
--bowtie2_genome BOWTIE2_GENOME
Bowtie2-indexed genome file. (default: None)
--fastq_r1 FASTQ_R1 Input fastq r1 file. Reads in this file are primed
from the provided primer sequence (default: None)
--fastq_r2 FASTQ_R2 Input fastq r2 file (default: None)
--fastq_umi FASTQ_UMI
Input fastq umi file (default: None)
--novel_cut_merge_distance NOVEL_CUT_MERGE_DISTANCE
Novel cut sites discovered within this distance (bp)
from each other (and not within
known_cut_merge_distance to a known/provided cut site
or a site with homology to guide_sequences) will be
merged into a single cut site. Variation in the cut
sites or in the fragments produced may produce
clusters of cut sites in a certain region. This
parameter will merge novel cut sites within this
distance into a single cut site. (default: 50)
--known_cut_merge_distance KNOWN_CUT_MERGE_DISTANCE
Novel cut sites discovered within this distance (bp)
with a known/provided/homologous site (that is not the
origin) will be merged to that site. Homologous sites
are defined as those that have homology to
guide_sequences. Novel cut sites farther than
known_cut_merge_distance will be merged into novel cut
sites based on the parameter novel_cut_merge_distance.
(default: 50)
--origin_cut_merge_distance ORIGIN_CUT_MERGE_DISTANCE
Reads aligned within this distance (bp) to the origin
site will be merged to that origin. (default: 10000)
--short_indel_length_cutoff SHORT_INDEL_LENGTH_CUTOFF
For reads aligned to a cut site, indels this size or
shorter are classified as "short indels" while indels
larger than this size are classified as "long indels"
(default: 50)
--suppress_homology_detection
If set, detection of guide sequence homology at cut
sites is skipped. By default, novel cut sites are
checked for homology, which can be computationally
demanding if there are many cut sites. (default:
False)
--PAM PAM PAM for in-silico off-target search (default: None)
--casoffinder_num_mismatches CASOFFINDER_NUM_MISMATCHES
If greater than zero, the number of Cas-OFFinder
mismatches for in-silico off-target search. If this
value is zero, Cas-OFFinder is not run (default: 0)
--primer_seq PRIMER_SEQ
Sequence of primer (default: None)
--primer_in_r2 If true, the primer is in R2. By default, the primer
is required to be present in R1.
(default: False)
--min_primer_aln_score MIN_PRIMER_ALN_SCORE
Minimum primer/origin alignment score for trimming.
(default: 40)
--min_primer_length MIN_PRIMER_LENGTH
Minimum length of sequence required to match between
the primer/origin and read sequence (default: 30)
--min_read_length MIN_READ_LENGTH
Minimum length of read after all filtering (default:
30)
--transposase_adapter_seq TRANSPOSASE_ADAPTER_SEQ
Transposase adapter sequence to be trimmed from reads
(default: CTGTCTCTTATACACATCTGACGCTGCCGACGA)
--arm_min_matched_start_bases ARM_MIN_MATCHED_START_BASES
Number of bases that are required to be matching (no
indels or mismatches) at the beginning of the read on
each "side" of the alignment. E.g. if
arm_min_matched_start_bases is set to 5, the first and
last 5bp of the read alignment would have to match
exactly to the aligned location. (default: 10)
--arm_max_clipped_bases ARM_MAX_CLIPPED_BASES
Maximum number of clipped bases at the beginning of
the alignment. Bowtie2 alignment marks reads on the
beginning or end of the read as "clipped" if they do
not align to the genome. This could arise from CRISPR-
induced insertions, or bad alignments. We would expect
to see clipped bases only on one side. This parameter
sets the threshold for clipped bases on both sides of
the read. E.g. if arm_max_clipped_bases is 0, read
alignments with more than 0bp on the right AND left
side of the alignment would be discarded. An alignment
with 5bp clipped on the left and 0bp clipped on the
right would be accepted. An alignment with 5bp clipped
on the left and 3bp clipped on the right would be
discarded. (default: 0)
--ignore_n If set, "N" bases will be ignored. By default (False)
N bases will count as mismatches in the number of
bases required to match at each arm/side of the read
(default: False)
--suppress_poor_alignment_filter
If set, reads with poor alignment (fewer than
--arm_min_matched_start_bases matches at the
alignment ends or more than --arm_max_clipped_bases on
both sides of the read) are included in final
analysis and counts (default: False)
--crispresso_min_count CRISPRESSO_MIN_COUNT
Min number of reads required to be seen at a site for
it to be analyzed by CRISPResso (default: 50)
--crispresso_max_indel_size CRISPRESSO_MAX_INDEL_SIZE
Maximum length of indel (as determined by genome
alignment) for a read to be analyzed by CRISPResso.
Reads with indels longer than this length will not be
analyzed by CRISPResso, but the indel length will be
reported elsewhere. (default: 50)
--crispresso_min_aln_score CRISPRESSO_MIN_ALN_SCORE
Min alignment score to reference sequence for
quantification by CRISPResso (default: 20)
--crispresso_quant_window_size CRISPRESSO_QUANT_WINDOW_SIZE
Number of bp on each side of a cut to consider for
edits (default: 1)
--run_crispresso_on_novel_sites
If set, CRISPResso analysis will be performed on novel
cut sites. If false, CRISPResso analysis will only be
performed on user-provided on- and off-targets
(default: False)
--cutadapt_command CUTADAPT_COMMAND
Command to run cutadapt (default: cutadapt)
--samtools_command SAMTOOLS_COMMAND
Command to run samtools (default: samtools)
--bowtie2_command BOWTIE2_COMMAND
Command to run bowtie2 (default: bowtie2)
--crispresso_command CRISPRESSO_COMMAND
Command to run crispresso (default: CRISPResso)
--casoffinder_command CASOFFINDER_COMMAND
Command to run casoffinder (default: cas-offinder)
--n_processes N_PROCESSES
Number of processes to run on (may be set to "max")
(default: 1)
--dedup_input_on_UMI If set, input reads will be deduplicated based on UMI
before alignment. Note that if this flag is set
deduplication by alignment position will be redundant
(only one read will exist with a UMI after this step).
This will also affect the values in the column
"reads_with_same_umi_pos" in the final_assignments.txt
file, which will only show 1 for all reads. (default:
False)
--suppress_dedup_on_aln_pos_and_UMI_filter
If set, reads that are called as deduplicates based on
alignment position and UMI will be included in final
analysis and counts. By default, these reads are
excluded. (default: False)
--dedup_by_final_cut_assignment_and_UMI
If set, deduplicates based on final cut assignment -
so that reads with the same UMI with different
start/stop alignment positions will be deduplicated if
they are assigned to the same final cut position
(default: False)
--umi_regex UMI_REGEX
String specifying regex that UMI must match (e.g
NNWNNWNNN) (default: None)
--min_umi_seen_to_keep_read MIN_UMI_SEEN_TO_KEEP_READ
Minimum times a UMI/read combination must be seen in
order to keep that for downstream analysis. If many
PCR cycles are performed in library preparation,
UMI/read combinations that are highly amplified may be
more trusted than UMI/read combinations that appear in
low abundance. However, this probably only applies for
sequencing libraries with members with uniform PCR
amplification properties. (default: 0)
--write_UMI_counts If set, a file will be produced containing each UMI
and the number of reads that were associated with that
UMI (default: False)
--r1_r2_support_max_distance R1_R2_SUPPORT_MAX_DISTANCE
Max distance between r1 and r2 for the read pair to be
classified as "supported" by r2 (default: 10000)
--suppress_r2_support_filter
If set, reads without r2 support will be included
in final analysis and counts. By default these reads
are excluded (default: False)