Differentiate editing efficiency of 2 guides used simultaneously

[JennyKou]

Hello, I am using Crispresso2 to quantify the % of modified reads in samples treated with a construct harboring 2 gRNA cassettes. I included both in my script (one as flexiguide, cause it has 2 mismatches with the target sequence). The % of modified reads contains reads with modifications on either target site or both.
Is there a way to quantify the % of reads that are considered modified for one or the other guide, apart from running again the script and supplying only one gRNA?

Thanks in advance for any help.

[kclem]

Hi @JennyKou,

Yes, if you want to quantify edits at only one guide, you could run each guide separately.

Alternately, for a detailed analysis by guide, please run the command with the --write_detailed_allele_table parameter. This will enhance your Alleles_frequency_table.zip file with the following columns:

• all_insertion_positions: This array specifies indices surrounding insertions, including those outside of the quantification window. For instance, [20, 21, 30, 31] indicates the presence of two insertions, one between the 20th and 21st base and another between the 30th and 31st base in the reference.

• all_insertion_left_positions: This array specifies indices of the left base for each insertion, encompassing those outside of the quantification window. In the example mentioned earlier, this value would be [20, 30].

• insertion_positions: This array specifies indices surrounding insertions, focusing solely on insertions that overlap with the quantification window. For instance, [20, 21] signifies one insertion that falls within the quantification window, between the 20th and 21st base in the reference.

• insertion_coordinates: This array specifies indices of insertions that overlap with the quantification window, presented as a tuple of (start, end), e.g., [(20, 21)].

• insertion_sizes: This array specifies insertion sizes corresponding to the insertions reported in insertion_coordinates. For example, [5] would indicate that the insertion between bases 20 and 21 was 5 base pairs long.

• all_deletion_positions: This array specifies indices of deleted bases, including those outside of the quantification window. For instance, [18, 19, 20, 50, 51, 52] suggests deletions of the 18-20th and 50-52nd bases in the reference.

• deletion_positions: This array specifies indices of deletions, considering only deletions that overlap with the quantification window. For example, [18, 19, 20] signifies a single deletion that intersects with the quantification window, removing the 20th and 21st base in the reference.

• deletion_coordinates: This array specifies indices of deletions overlapping the quantification window, presented as a tuple of (start, end), e.g., [(18, 21)].

• deletion_sizes: This array specifies deletion sizes corresponding to deletions reported in deletion_coordinates. For example, [3] would indicate that the deletion between bases 18 and 21 was 3 base pairs long.

• all_substitution_positions: This array specifies indices of substituted bases, encompassing those outside of the quantification window. For example, [20, 30, 35] implies substitutions at the 20th, 30th, and 35th bases in the reference.

• substitution_positions: This array specifies indices of substitutions, considering only those that overlap with the quantification window. For example, [20] indicates a single substitution within the quantification window at the 20th base in the reference.

• substitution_values: This array corresponds to substitution_positions and shows the resulting base after substitution. For instance, ['A'] means that the 20th base was substituted, resulting in an 'A'.

Generally, the column names that begin with all_ refer to all sequence changes, including those outside of the quantification window, whereas other columns refer only to sequence changes that overlap the quantification window. Thus, all_insertion_positions will contain all insertions in the alignment, whereas insertion_positions will only include insertions that overlap with the quantification window. Thus, insertion_positions will always be a subset of all_insertion_positions.

Using these additional columns, you can identify which sequences contained an edit at one or multiple quantification windows.

Let me know if this make sense!

[JennyKou]

Great thanks for the tip! For now I am happy with just the % of modified reads for each guide. At the beginning I tried using the Alleles_frequency_table_around_GuideX.txts to differentiate between each guide, but I don't think they gave me the right answer.

[kclem]

I just wrote a little script to do this parsing that will work with version 2.2.15 or greater a1435f7.

I don't have a good double-edited sample handy, but it can be run on the demo HDR data hdr.fastq.gz using the command:

CRISPResso -r1 hdr.fastq.gz -a acatttgcttctgacacaactgtgttcactagcaacctcaaacagacaccatggtgcatctgactcctgTggagaagtctgccgttactgccctgtggggcaaggtgaacgtggatgaagttggtggtgaggccctgggcaggttggtatcaaggtta -e acatttgcttctgacacaactgtgttcactagcaacctcaaacagacaccatggtgcaCctgactccGgaggagaagtctgccgttactgcGctgtggggcaaggtgaacgtggatgaagttggtggtgaggccctgggcaggttggtatcaaggtta -c atggtgcatctgactcctgTggagaagtctgccgttactgccctgtggggcaaggtgaacgtggatgaagttggtggtgaggccctgggcag -g TGCACCATGGTGTCTGTTTG,GATGAAGTTGGTGGTGAGGCCC --write_detailed_allele_table  -n hdr3 -p max -gn guide1,guide2

python CRISPResso2/scripts/count_sgRNA_specific_edits.py -f CRISPResso_on_hdr3

This produces:

Processed 25000 alleles
Reference: Reference (2391/23415 modified reads)
        UNMODIFIED: 21024
        MODIFIED guide1: 2359
        MODIFIED guide2: 32
Reference: HDR (856/1577 modified reads)
        UNMODIFIED: 721
        MODIFIED guide1: 854
        MODIFIED guide1 + guide2: 1
        MODIFIED guide2: 1