CRISPResso - Alignment with inserts rather with edited alleles by base editor

[ulinhho]

Hi everyone,

I have an issue with the CRISPResso analysis of some of my NGS samples (Input: paired end fastq files, amplicon sequence and gRNA added).

The goal is to see the editing efficiency of the base editor and obtain the alleles with the different edits. I tried CRISPResso through conda and also from the web interface. CRISPResso --fastq_r1 37_Det_2_R1.fastq.gz --fastq_r2 37_Det_2_R2.fastq.gz --amplicon_seq TGGCTGGCTTTGATCTGGACGGGACGCTCATCACCACACGCTCTGGGAAGGTCTTTCCCACTGGCCCCAGTGACTGGAGGTGATAAGAGGC --guide_seq ACGCTCATCACCACACGCTC --quantification_window_size 10 --quantification_window_center -10 --base_editor_output done For the web interface, I also stated to ignore insertions, and deletions.

I expected CRISPResso2 to align the amplicons with base edits (point mutations as I used a Cytidine deaminator).

For both, the software plotted only the small amount of inserts but not the sequences with the edits. I had a look with the Integrative Genomics Viewer (IGV) from the Broad Institute and it seems that IGV could detect 20295 reads (also the amount of reads I expect to see) with high base editing efficiencies (see picture).

As IGV can only give me positional editing, I would want to make the analysis in CRISPResso2 work as it gives me more information about the alleles. Can you help me to fix the bug?

Thanks in advance!

Debug output Paste the entire out CRISPResso_RUNNING_LOG.txt 1a.Read_barplot.pdf put when you run CRISPResso with the flag --debug.

[kclem]

Hi @ulinhho sorry about that - I would expect that CRISPResso should work in this case.

From your output it appears that CRISPResso cannot align your reads to the reference sequence. By default it requires at least 60% homology between the read and the reference sequence (a cutoff set by the --default_min_aln_score parameter). I'm not sure exactly what is going on, but it appears that your reference sequence is 91bp long. How long were your sequencing reads? If they were longer than 91bp you're going to have sequencing adapter readthrough, which won't align to your reference sequence, and if fewer than 60% of bases match your reference sequence the alignment will fail.

Can you try the following: 1) try trimming sequencing adapters from your R1 and R2 files before inputting them to CRISPResso 2) Run flash on your R1/R2 files before CRISPResso. Check the output of flash to make sure your expected amplicon appears and that there are no sequencing adapters. You can also run fastp on your reads, which will attempt to remove adapter readthrough from your reads when merging R1/R2. After you have merged R1/R2, run the merged reads into CRISPResso. 3) set --default_min_aln_score to 0: CRISPResso --fastq_r1 37_Det_2_R1.fastq.gz --fastq_r2 37_Det_2_R2.fastq.gz --amplicon_seq TGGCTGGCTTTGATCTGGACGGGACGCTCATCACCACACGCTCTGGGAAGGTCTTTCCCACTGGCCCCAGTGACTGGAGGTGATAAGAGGC --guide_seq ACGCTCATCACCACACGCTC --quantification_window_size 10 --quantification_window_center -10 --base_editor_output --default_min_aln_score 0 This should align reads to your amplicon even if they have a lot of insertions (sequencing adapter) at the end. 4) Run with --auto mode to discover amplicons that are present in your reads file.

Let me know if those don't work and I can take another look.

[ulinhho]

Thanks for the help! The sequencing reads with all the flow cell attachment sequences, Illumina sequencing sequences, i5/i7 indexes, spacers are between 270 and 300 bp. One thing I forgot to mention is that CRISPResso works for most of the amplicons (10 out of 12, same adaptors too). Only for two amplicons it did not work. I will try your suggestions out and let you know if I can fix the problem.

[kclem]

Hi @ulinhho, I hope this worked for you, please reopen this issue if you are still having problems!