pinellolab / CRISPResso2

Analysis of deep sequencing data for rapid and intuitive interpretation of genome editing experiments
Other
270 stars 94 forks source link

Prime Editing Issue: Unexpected sgRNA Cutting within Amplicon #368

Closed CharlotteMarasigan closed 8 months ago

CharlotteMarasigan commented 8 months ago

Hello Community,

I am writing to seek your expertise regarding an unexpected issue I encountered while attempting to generate Prime Editing results. I meticulously followed the CRISPRESSO manual and inputted all required information for targeting the desired amplicon. However, upon analysis, I found that the sgRNA is cutting within the amplicon, specifically at the middle when it is not supposed to be, instead of at the intended editing site.

I have already tried the troubleshooting steps provided on the website to address this issue but haven't been successful. I would be incredibly grateful if you could provide any insights or suggestions on how to resolve this discrepancy. I have also attached some documents for your reference. Thank you very much.

Regards,

Charlotte Marasigan Explanations .docx Raw Reads Results .zip

kclem commented 8 months ago

Hi @CharlotteMarasigan ,

Sorry you're having troubles with this analysis. It appears that your pegRNA binds to several locations within the amplicon sequence. Is this correct?

CGATGCAGACGATGCAGACGATGCAGACGATGCAGACGATGCAGG < reference amplicon
   TGCAGACGATGCAGACGATGC < spacer site 1
            TGCAGACGATGCAGACGATGC < spacer site 2
                     TGCAGACGATGCAGACGATGC < spacer site 3

When running in Prime Editing mode, CRISPResso creates a new reference amplicon called 'Prime Edited' and aligns reads to that amplicon. However, in your case, multiple 'Prime Edited' amplicons are possible, so the analysis becomes much more complicated.

Are you trying to analyze prime editing at multiple of these sites? Or are you only interested in a single site? Are you interested in scaffold incorporation? Or only determining which sites are edited?

For this example, I would probably input all expected outcomes as separate amplicons and run in normal CRISPResso mode (not prime editing) to figure out counts for each prime editing outcome. Feel free to reach out at k.clement@utah.edu if you'd like to talk about how best to achieve your analysis results.