Just as a check for the substitutions I was observing, I reversed R1.fastq and R2.fastq being the paired end reads from as sequencing experiment. Let's say at 'chr1 100' that I observe and 'A -> G' substitution. My thought here was that if I switch the paired end files so that R1.fastq becomes R2.fastq and then align and call variants, that my base substitution would now be the complimentary 'T -> C' change (using bwa mem to align and freebayes to call vars). Yet when I do this, I still get and 'A->G' change.
Am I observing no change in the substitution because of the processing by freebayes and bwa mem? Do they align to either strand and then always report the substitution on the positive strand?