When a large number of living cells are targeted simultaneously using CRISPR, the use of bioinformatic tools can speed up the process of identifying the sequence variations of CRISPR products from targeted indels, insertion, or deletion. Amplifying the targeted DNA using PCR followed by cloning the sequence variations into plasmids and sequencing them one by one produces accurate results but are time consuming. Recently I've been looking into several online bioinformatic tools such as TIDE/TIDER and ICE, and although these tools are useful for detecting indel variations, their developers stated that they are not suitable for detecting large insertions/deletions. Has anyone tried using these tools to detect large insertions/deletions? Any recommendation for a different tool/strategy to identify the sequence variations of pooled CRISPR products?
Hi Andika. It is common to use Next Generation sequencing to analyse what happens after CRISPR editing in the pool, like Illumina or others. Illumina double-end sequencing will give you up to 500-600 bp long reads. So you should be fine for 200 + but, not 400+ . One can move NGS sequencing primers around your target to capture larger deletions/insertions. My understanding that some other NGS reads are even longer, so may be that the way to go. That what concerns -PCR. In terms ao the software - I am not sure, but you could analyse what happened even in Excel. Sequence output from NGSis sortable and countable.
Hello Andika. I have no experience with the tools you mentioned. However, an excellent bioinformatic tool which is suitable for the detection of large indels (using short reads) is the broadinstitute software tool Pilon available at the link below.
https://github.com/broadinstitute/pilon/wiki
In my experience it performed very well in detecting large insertions and deletions (>1.5 Kb) from whole genome sequencing Illumina pair-end reads (150 bp) of rearranged genomes (product of Programmed DNA Elimination in my case rather than purposed genome editing) . I suspect that the variants in your pooled CRISPR products should be sufficiently represented in the pool for the variant calling to be successful. Perhaps it's worth a shot.
Best,
A "quick and dirty" method to just screen your data could be to simply plot the read depth (on the y axis) over the genomic position (on the x axis) of two sets ("control" vs CRISPR transformant) at a time. In theory, your inserts will have a higher read depth vs. your control and you could easily visualize these in the plots. Of course, this in silico analysis would not spare you the work of actually confirming the inserts in vitro (if you want to publish the work), but it can help you screen large data sets...
- Badges
- Science topic
- Similar topics
- Computer Science and Engineering
- Bioinformatics
- Bioinformatic Tools