What's the best short read aligner now?

[gradient]

I've been out of the loop for a bit and I'm wondering what is considered the best for aligning reads around 50bp-300bp to human genome. By best, I am referring to fastest runtime with above average quality. FPGA and GPU aligners would be interesting too.

Thanks.

[GenoMax]

Don't get people started

Earlier this year in a Biostars thread: https://www.biostars.org/p/125020/

You have to try BBMap out after reading that thread though.

[blancha]

My vote is for Bowtie2, on the type of reads you described.
I use nearly exclusively Bowtie2, without ever having experienced any issues.
For Exome-Seq only, I use BWA since it is the aligner used in the Broad Institute Best Practices.
I'm pretty sure Bowtie2 would work just as well though.
I seem to remember having use Bowtie2 once by accident on Exome-Seq data, and found to my surprise, that the results were pretty much the same when I repeated the alignment with BWA, give or take a few reads.

I've come to the general conclusion that the results for the main aligners are so similar for the type of alignments that you are describing, that it is irrelevant which aligner is used.
I've tried BWA, Bowtie2, and Bowtie on the same datasets with near identical results.
I didn't keep the results, so I couldn't show them to you right now.

I'm quite skeptical of benchmarks, since they never seem to be available for the current versions of the programs. There have been more than seven releases of Bowtie2 since the benchmarks on the website given above were published.

My only advice would be to pick any of the main aligners and become very familiar with its parameters and the format of the output.
There are some variations in the default settings, the parameters that can be changed, and the format of the output.

[baseq]

Regarding your interest in GPU aligners, you might check out SOAP3:

http://soap.genomics.org.cn/soap3.html

http://journals.plos.org/plosone/article?id=10.1371/journal.pone.0065632

I included the journal article for its table comparing performance of many of the aligners you'll read about.

Good luck!