samtools flagstat mismatch in stats

[nilshomer]

When the flag field is '*', does it mean an unmapped read? I know that a flag 4 is definitely unmapped.

The '*' in the flag field is invalid, and does not conform to the specification (see the "Note" in Section 2.2.1).

[bioinfosm]

When the flag field is '*', does it mean an unmapped read? I know that a flag 4 is definitely unmapped.

[nilshomer]

Hi all,
I have a bam file created from illumina alignments with bwa. When I try to get the number of aligned reads 2 different ways I get 2 different numbers...

When I run flagstat:
samtools flagstat myBam.bam
I get this many mapped reads
41721001 mapped (93.82%)

However when I try to get this on my own with this command:
samtools view myBam.bam | awk -F '\t' '($3 != "*")' | wc -l
I get this number
42283661

I'm sure I'm doing something wrong, but not sure what. Is it incorrect to think that all reads with a non '*' listed in the third column should be considered aligned?

thanks for the help!

You want:

samtools view -f 4 myBam.bam | wc -l

See the FLAG field in the SAM spec.

[rcorbett]

Hi all,
I have a bam file created from illumina alignments with bwa. When I try to get the number of aligned reads 2 different ways I get 2 different numbers...

When I run flagstat:
samtools flagstat myBam.bam
I get this many mapped reads
41721001 mapped (93.82%)

However when I try to get this on my own with this command:
samtools view myBam.bam | awk -F '\t' '($3 != "*")' | wc -l
I get this number
42283661

I'm sure I'm doing something wrong, but not sure what. Is it incorrect to think that all reads with a non '*' listed in the third column should be considered aligned?

thanks for the help!