BWA and paired end reads

[scami]

Hi guys
I am aligning some paired end reads on a reference genome using the "bwa sampe" command. The resequenced organism is the same species and ecotype of the one used to build up the reference genome and for this reason I was expecting a high coverage. I did not get a coverage which was higher than the one obtained by aligning different ecotypes and something else really surprised me: if I increase the edit distance I am supposed to align more reads on the reference genome. However I get the opposite behavior. I get the maximum coverage using the minimum edit distance (n=1). I was wondering: when a read map in multiple positions on the reference genome are they all reported in the .sai file which is generated by running the bwa aln command? When the bwa sampe command place the pared end reads on the reference genome does it do all the possible combinations between the multiple positions which are found on the sai files.
thanks for the help!

[swbarnes2]

When the bwa sampe command place the pared end reads on the reference genome does it do all the possible combinations between the multiple positions which are found on the sai files.
thanks for the help!

No. Bwa will pick one position to report per read. It will try to select a mapping so that the pair in question has an insert size close to that observed in other pairs, but there will only be a single .sam line per read.

[scami]

thanks!

Thank you very much for your answer. I appreciated it. Still I can not understand why if I increase the edit distance I get a lower average coverage. Isn't is unusual?