0% Alignment Rate with Bowtie2 in CRISPRseq Workflow

[vikramadhithyaThotam22]

Description: I am running a CRISPR screening workflow using Bowtie2 for alignment and MAGeCK for counting. My input is an sgRNA FASTQ file, and my reference was initially in CSV format, which I converted into FASTA before starting the workflow.

However, I encountered an issue where all reads failed to align according to the Bowtie2 log:

2788818 reads; of these: 2788818 (100.00%) were unpaired; of these: 2788818 (100.00%) aligned 0 times 0 (0.00%) aligned exactly 1 time 0 (0.00%) aligned >1 times 0.00% overall alignment rate

Despite this, I obtained a BAM file. I want to understand why the alignment rate is 0% and how to obtain a correctly aligned BAM file for MAGeCK.

For reference, here is my expected alignment summary from MAGeCK count: 2P241219012UG192662DX_AND-RTR-LIB-22_L01.fastq-2788818 total reads-2558289 Mapped-0.9173 percentage

Any insights or troubleshooting suggestions would be greatly appreciated!

[matbonfanti]

Hi,

I do not know if this is the case, but I had a similar issue lately. My problem was that after trimming I was left with sequences that were much longer than the 20 bases that are the actual length of the guide. With the default bowtie parameters, the part of the read that do not match the guide lowers the alignment score below the minimum for alignment. The solution I came up with is to change the trimming parameters and to keep only 20 bases (the length of the guide) after the adapter. One could also change the parameters of bowtie to allow soft-clipping of the bases of the read beyond the guide, as an example.

Hope this helps!