CRISPRessoBatchPooled?

[cdustinr]

Hi @kclem,

As always, thanks for your work on this amazing resource.

I often run many samples (dozens to hundreds) and find CRISPRessoBatch very appealing. But often these libraries consist of pooled amplicons (completely unrelated loci, and pooled post-PCR). For individual samples with pooled amplicons, CRISPRessoPooled is very effective.

As far as I can tell, it isn't possible to run CRISPRessoBatch in Pooled mode (though if I'm wrong that would be wonderful). How would you recommend configuring these analyses?

If it may be relevant, sometimes these amplicons sequences are very similar (i.e., gene & pseudogene), so allowing each read to map to the best "region" is preferred to running each "region" (i.e., gene & pseudogene) sequentially. This can be addressed with --amplicon_min_alignment_score, but large indels & highly similar sequences could be problematic.

Thanks again!
Dustin

[kclem]

Thanks @cdustinr!

From what I understand, you're running CRISPRessoPooled for lots of different amplicons. You'd like to aggregate information for amplicons across CRISPRessoPooled runs, like CRISPRessoBatch. Is that correct?

Have you checked out the CRISPRessoAggregate program? If you provide your CRISPRessoPooled folders to this program, it should go through and aggregate editing details by amplicon. Is that what you're after?

Thanks,

Kendell

[cdustinr]

Thanks @kclem! Yeah, I think that would do the trick. Tbh, I thought CRISPRessoAggregate was obligately used by CRISPRessoPooled, and I didn't realize CRISPRessoAggregate was exposed and callable by itself.

So is running CRISPRessoBatch equivalent to running CRISPResso + CRISPRessoAggregate? And I suppose CRISPRessoAggregate can handle the subfolders created by CRISPRessoPooled?

Thanks again!
Dustin

[kclem]

The aggregate steps of CRISPRessoBatch and CRISPRessoPooled are similar to CRISPRessoAggregate, but I pulled pulled out the aggregation to a separate program for users like you who had more complex experimental designs.

Yes, it should be able to pull individual CRISPResso runs inside a CRISPRessoPooled run, and aggregate those by amplicon sequence. Let me know if it works for you the way you expect, or if you have any suggestions!

[cdustinr]

This is great. Thanks @kclem!