search

lucapinello / CRISPResso

Software pipeline for the analysis of CRISPR-Cas9 genome editing outcomes from sequencing data
Other
131 stars 55 forks source link

Appears to be counting reads with deletions as unedited #50

Closed AGRaeside closed 5 years ago

AGRaeside commented 5 years ago

Hi Luca, Big fan of crispresso and very much looking forward to crispresso 2. I was running a batch of crispressos and I came across a few wells that seems to have a usually high amount of unmodified reads. When i looked at the alleles around the predicted cut side they appeared to have clear deletions but were counted as unmodified in every statistic.

Aligned_Sequence Reference_Sequence Unedited %Reads #Reads AGTGGAGGATGCCTTCT--ACGTTGGTGCGTGAGATCCGG AGTGGAGGATGCCTTCTACACGTTGGTGCGTGAGATCCGG True 61.50844322453524 103082 GATGCCTTC-ACATGTCTCACGTTGGTGCGTGAGATCCGG GATGCCTTCTA-------CACGTTGGTGCGTGAGATCCGG True 35.92815800465359 60212 AGTGGAGGATGCCTTCT----------------------- AGTGGAGGATGCCTTCTACACGTTGGTGCGTGAGATCCGG False 0.6533802732859946 1095

I'm not sure why it CRISPresso could be counting these reads as unmodified. I think there is the same thing in the reads in the left-aligning option issue where the reads have deletions but are counted as unedited.

Any help would be deeply appreciated.

Thanks,

Alexander Raeside Oxford Genetics

lucapinello commented 5 years ago

This depends on the window size you have set around the cleavage site. If those indels are not overlapping the window they will not be counted.