Read alignment using Bowtie2

Question

[Ina]

Hello everyone!

I am in need of using Bowtie2 to find all possible alignments for a set of reads (each 23bp long) to the human genome. The maximum number of mismatches that can be tolerated is 6. But since Bowtie2 allows mismatches only upto 1, I had to set the mismatch penalty for alignment score accordingly to get the result. Since my laptop does not have the capacity to perform this operation, I am searching for interfaces that provide access to Bowtie2. I found Galaxy to be useful but unfortunately, I cannot see an option to set the alignment score in it. Does anyone know of any other tool that provides an option for Bowtie2 with alignment score?

Thanks!

Answer 1

[ATpoint]

For ultra-short reads it is recommended to use bowtie (aka bowtie1) which allows you to explicitely set the number of tolerated mismatches. From the bowtie2 FAQs:

For reads longer than about 50 bp Bowtie 2 is generally faster, more sensitive, and uses less memory than Bowtie 1. For relatively short reads (e.g. less than 50 bp) Bowtie 1 is sometimes faster and/or more sensitive.

http://bowtie-bio.sourceforge.net/bowtie2/faq.shtml

Be aware that with this short length you'll get plenty of multimappers especially if allowing that many mismatches which is 26% of the total read length. What experiment is this, maybe there are better alternatives?

[Ina]

Hi,

Thank you ATpoint for your suggestion.

So this is related to CRISPR-CAS9. I am working with off-target predictions for my thesis and was looking at all scientific papers related to CRISPR. I found one and decided to use their datasets. The datasets were not available directly and the method used to obtain them was specified which I had mentioned below:

They have around 30 sgRNAs which are each 23bp long and are aligning the sgRNAs to human reference genome with a tolerance of upto 6 mismatches using Bowtie2.

In order to access the exact dataset, I thought I need to replicate the method they used. To do so, I collected the sgRNAs from their paper and tried to use Bowtie2 to align them against human genome with upto 6 mismatches. I decided that only Bowtie2 can be used since the output from different tools may not be deterministic.

[Friederike]

The authors may have misnamed the tool; perhaps they were working with Bowtie. However, unless they provide the specific command in the manuscript, there is no way of verifying this unless by sending the authors of the original work a message asking to clarify the details (which is the course of action I would suggest, even if this comes as a belated advise)