Is there an easy way to filter SAM/BAM files with NM>0

[dontkme]

Hi, everyone

BAM files contain the informations like
"AS:i:-8 XM:i:0 XO:i:0 XG:i:0 MD:Z:100 NM:i:0 XS:A:+ NH:i:1".
NM means not matched bp numbers.

Now I wanted to pick the reads that are totally mapped to the ref sequence.

I have tried to use Picard SamToFastq and tophat/bowtie2 -N 0 to get the final BAM, but samtools view -f 8 cannot correctly pick the single mapped reads, so SamToFastq didin't work well.

So is there an easy way to filter SAM/BAM files with NM>0 .

Thanks.

[lindenb]

Using my program sam.js https://github.com/lindenb/jvarkit/wiki/SamJS

 java -jar  dist/samjs.jar  -e 'parseInt(record.getAttribute("NM"))===0'  your.bam

[Chipper]

This should give you the NM:0 alignments in sam format:

samtools view file.bam | awk '$0 ~"NM:i:0"' > perfect-match.sam

[dontkme]

This should give you the NM:0 alignments in sam format:

samtools view file.bam | awk '$0 ~"NM:i:0"' > perfect-match.sam

Thanks it worked well.

I added the header and IGV can use the bam.