Hello ,

Thanks a lot for this much of help. Actually i am working with Human genome and problem infront of me is that when i run the BWA-MEM algorithm i get some unexpected results everytime.

Let me explain here

My read 1 and read 2 files have equal number of reads

Step followed:

Step 1 : bwa-mem run

Step 2: samtools flagstat command for statistics of SAM/BAM generated.

Step 3: Problem phase : flagstat command shows that read 1 and read 2 have unequal number of reads but before BWA-MEM run i checked twice that my files have equal number of reads.

I am attaching a screenhot regarding my problem

Please if you understand the problem then try to solve it for me.

I think it is because your read is mapping multiple positions in the genome. For example left read is mapping at two positions while the right read is mapping at only one position. It is very possible because a genome can have paralogs, but these paralogs are differing from each other for the second read.

BWA MEM algorithm is more powerful then the previous version. The great advantage is local alignment based on seed then extending it for full read. this process gives the possibility to obtain the optimal solution to mapping problem and hence decreasing false positives.

Hello A. Chaturvedi.

Thanks for responding and i checked as well. It helped me lot. But still i have one confusion left. Why read 1 and read 2 are unequal in number acc to flagstat command (see the attachment)??

Have a look at this thread:

http://seqanswers.com/forums/showthread.php?t=32708

Best,

Anurag