I'm now analysising targeted sequencing pair-end reads (approximately 100M per sample) to find edit pattern of crispr.
But how to do mapping and alignment? bowtie2 --very-sensitive-local (or -D 20 -R 3 -N 1 -L 20 -i S,1,0.50)or bwa-mem? which should be chosen as mapping tool.
Pro.Isvart say that bwa-mem is usually better than bowtie2(default). but when it comes to bowtie2 -D 20 -R 3 -N 1 -L 20 -i S,1,0.50. so question 1, what is the difference therotically?
I apply the two tools to my sample reads. the mapped reads number are nearly the same. but many reads in bwa sam files are secondary alignment (with FLAG 512)and some of them are not mapped in proper pairs(without FLAG 2),thus bwa may give less edited patterns.
so question 2, when a read is secondary alignment , does it appear more than once in sam file? question 3, if reads pair (read1 and read2) are not mapped in proper pairs, can these reads be used to count the pattern? question 4, can these flag make bwa more strict? Is bowtie2 results reliable?