Add bowtie2 as choice for aligner

[ivokwee]

Hi. I succesfully tried bowtie2 as aligner by installing bowtie2 in the docker image, then replacing the aligner code in main.nf with the code below. Everything seems to work fine but of course all results are in the results/bwa folder. I wonder if it would be nice to let the user choose between BWA-mem, or Bowtie2. Ivo

"""
bowtie2 \\
   -p $task.cpus \\
   -x ${index}/${bwa_base} \\
   -U $reads \\
   | samtools view -@ $task.cpus -b -h -F 0x0100 -O BAM -o ${prefix}.bam -
"""

[drpatelh]

Thanks @ivokwee. I was planning on adding in such options when this pipeline is eventually ported to DSL 2. Im a bit confused as to how the pipeline worked with the command you sent though because that implies that bowtie2 worked with a bwa index? The pipeline would need to be changed in a few places to cater for multiple aligners.

bowtie2 \\
   -p $task.cpus \\
   -x ${index}/${bwa_base} \\

[drpatelh]

From the benchmarking I have seen to date it appears that bwa and bowtie2 give incredibly similar results. I realise it's personal preference and mine has always been bwa but I am wondering whether it makes more sense to add other alignment options instead e.g. STAR for speed.

[ivokwee]

1. I had a local index for Bowtie2 (from iGenomes). 2. The sample data are from https://angus.readthedocs.io/en/2019/chip-seq.html. I don't know why the bwa scores were so low. With the default settings there were no mapped reads reported so I tracked down it was the T parameter. 3. I just see Bowtie2 often used. Also the default "just worked" in this case. 4. STAR would be nice too. Regards. Ivo.

…

On Tue, Jun 9, 2020, 17:26 Harshil Patel ***@***.***> wrote: From the benchmarking I have seen to date it appears that bwa and bowtie2 give incredibly similar results. I realise it's personal preference and mine has always been bwa but I am wondering whether it makes more sense to add other alignment options instead e.g. STAR for speed. — You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub <#160 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AEVULWRELZJXKWVXIQQIR73RVZIBDANCNFSM4NZNBLOQ> .

[JoseEspinosa]

As suggested by @drpatelh in #233, the DSL2 implementation of the pipeline could have a sub-workflow that implements bowtie2 as an aligner, see here the current bwa_men implementation as a template.