Mapping bacterial genome with BWA - best seed length and algorithm?

[mrood]

Hi There,

I am mapping bacterial genomes to a reference sequence and am wondering if anyone has experience or knowledge with the best algorithm to use and/or any suggestion about seed length. (I'm new to bioinformatics!) Some of my samples are 95bp paired-end reads, and some are 115bp paired-end reads.

Any suggestions on optimal parameters to use will be greatly appreciated! (Note: I have trimmed the sequences.)

[mastal]

Mapping bacterial genome with BWA - best seed length and algorithm?

The popular short-read aligners BWA and Bowtie/Bowtie2 should work fine if you have a reference genome. Just use the default parameters in the first instance.

[oiiio]

@mrood

Check out these alignment benchmarks for some different popular tools (BWA, Bowtie2, Novoalign) here:

http://www.bioplanet.com/gcat/reports/23/alignment/100bp-pe-small-indel/bowtie/compare-18-22-16

As for parameters, default parameters should be fine with any of those. Personally, I would recommend BWA-MEM as it is fast and accurate for those read lengths.

[mrood]

Thank you mastal and oiiio!

I previously mapped with the bwa aln sampe and I have tried the bwa mem on one sample. From a quick comparison by eye in IGV it looks like the bwa mem will work better for my samples. I forgot to mention that they are pooled samples, so proper mapping is very critical for downstream analyses.

I appreciate both of your fast responses!