I've read about on-target efficiency and off-target effects in use of CRISPR/Cas9, and about tools that suggest good guide sequences. I am wondering: how many guide sequences do typical CRISPR users try -- i.e. is there any iteration or parallelism involved?

I think the answer is yes for some use cases, but I am not sure:

  • Knocking out one particular locus: There are only a few possible guide sequences, so you can just make a library that includes all of those and run one experiment, without iteration. You would not know which guide sequence worked best, though, but users may not need this information.
  • Knocking out one gene, no matter at which loci: Since there are many possible loci for indels, the key is to choose a locus for which the guide sequence is effective and produces the last off-target effects elsewhere in the genome. This would mean iterating/trying several of the top suggested guide sequences individually. These could be tried in parallel across several experiments.
  • Knocking out multiple genes, aka multiplexing: The user would want to find good guide sequences for each gene separately to avoid any interactions. This would reduce to the previous example above.

Is my understanding above correct?

  • 1
    $\begingroup$ I usually do 2-3 parallel guides for a single gene, but I'm not sure on how common it is. $\endgroup$
    – March Ho
    Jul 26 '16 at 3:37
  • $\begingroup$ @MarchHo Thanks! Do you just use the 2-3 top guides from a guide selection tool like crispr.mit.edu in parallel, and then report whichever worked best? $\endgroup$ Jul 26 '16 at 16:34

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.