[Bio-bwa-help] Need help with using BWA for short queries contain high error rate (up to 6%)

[Yangyi C. <che...@ya...>]

I need to map 100bp sequences with up to 6 errors (mismatches and indels) to 
human genome (hg19), however, I am a little bit confused about the optimal way 
to do it using BWA.

1. Which alignment mode should I use?
As mentioned in the manual of BWA, the aln mode can be used to deal with queries 
up to ~200bp with <3% error rate, bwa-sw usually is used to align longer reads 
(more than 350bp ??). For the dataset I used (short queries with high error 
rate), which mode should I choose? As for the results, I only care about whether 
the queries can be mapped to human genome or not within a given error rate 
(<6%), I don't care about all the locations a queries can be mapped to or the 
optimal location.  

2. How do I set the options for each mode?
Currently I am only adding options -n 6 -o 6 for aln mode and using the default 
options for bwa-sw mode. Should I change or add any other option? 

Thanks a lot!