Splitting a bam file by unique optional TAG field

Question

[dccroucher]

I have a large bam file that I need to subset/split into separate bam files by unique entries of one of the optional TAG:TYPE:VALUE fields. As a simplified example, say these are the optional fields in my example.bam file (i.e. fields following the mandatory 1-11 fields):

NH:i:6  HI:i:3  AS:i:94 nM:i:1  NM:i:1  CR:Z:TGCCCATCACTCAGGC   CY:Z:CCCBBFGGGGGGGGGG   CB:Z:TGCCCATCACTCAGGC-1  UR:Z:ACGCCCAGGT  UY:Z:GGGGGGGGGG  UB:Z:ACGCCCAGGT  BC:Z:TGCGCAGC   QT:Z:CCCCCGGG   RG:Z:XXXXXX

NH:i:6  HI:i:3  AS:i:94 nM:i:1  NM:i:1  CR:Z:TGCCCATCACTCAGGC   CY:Z:CCCBBFGGGGGGGGGG   CB:Z:TGCCCATCACTCAGGC-1 UR:Z:ACGCCCAGGT 

NH:i:7  HI:i:3  AS:i:96 nM:i:0  NM:i:0  CR:Z:TGCCCATCACTCAGGC   CY:Z:BCBBCFGGGGGGGGGG   CB:Z:TGCCCATCACTCAGGC-1 UR:Z:ACGCCCAGGT 

NH:i:7  HI:i:3  AS:i:96 nM:i:0  NM:i:0  CR:Z:TGCCCATCACTCAGGC   CY:Z:CCCCCGGGGGGGGGGG   CB:Z:TGCCCATCACTCAGGC-1 UR:Z:ACGCCCAGGT 

NH:i:3  HI:i:2  AS:i:96 nM:i:0  NM:i:0  CR:Z:TGCCCATCACTCAGGC   CY:Z:CCCCCGGGGGGGGGGG   CB:Z:TGCCCATCACTCAGGC-1 UR:Z:ACGCCCAGGT 

NH:i:3  HI:i:2  AS:i:96 nM:i:0  NM:i:0  CR:Z:TGCCCATCACTCAGGC   CY:Z:CCCCCGGGGGGGGGGG   CB:Z:TGCCCATCACTCAGGC-1 UR:Z:ACGCCCAGGT 

NH:i:6  HI:i:4  AS:i:96 nM:i:0  NM:i:0  CR:Z:CCGGTAGGTCATACTG   CY:Z:CCCCCGGGGGGGGGGG   CB:Z:CCGGTAGGTCATACTG-1 UR:Z:GCCGCCTTCT 

NH:i:6  HI:i:4  AS:i:96 nM:i:0  NM:i:0  CR:Z:CCGGTAGGTCATACTG   CY:Z:CCCCCGGGGGGGGGGG   CB:Z:CCGGTAGGTCATACTG-1 UR:Z:GCCGCCTTCT 

NH:i:6  HI:i:4  AS:i:96 nM:i:0  NM:i:0  CR:Z:CCGGTAGGTCATACTG   CY:Z:CCCCCGGGGGGGGGGG   CB:Z:CCGGTAGGTCATACTG-1 UR:Z:GCCGCCTTCT 

NH:i:6  HI:i:4  AS:i:96 nM:i:0  NM:i:0  CR:Z:CCGGTAGGTCATACTG   CY:Z:BBBCBGGEGGGGGGGG   CB:Z:CCGGTAGGTCATACTG-1 UR:Z:GCCGCCTTCT 

NH:i:6  HI:i:4  AS:i:96 nM:i:0  NM:i:0  CR:Z:CCGGTAGGTCATACTG   CY:Z:CCCCCGGGGGGGGGGG   CB:Z:CCGGTAGGTCATACTG-1 UR:Z:GCCGCCTTCT

How would I split this bam file by unique CB:Z: entries, such that the results would be 2 bam files as follows:

TGCCCATCACTCAGGC-1.bam (contains 6 reads) CCGGTAGGTCATACTG-1.bam (contains 5 reads)

Thanks! D

[GenoMax]

See this past thread: Sorting .bam file by tag
looks like bamtools (https://github.com/pezmaster31/bamtools/wiki ) may be able to do this.

[marong0511]

bamtools split -in test.bam -tag CB

Try this but the problem is there are too many files and bamtools can't open the bam file for writing.

See https://github.com/pezmaster31/bamtools/issues/135

Answer 1

[Pierre Lindenbaum]

run a loop to find the distinct tag and pipe this into a samtools+awk to filter. Something like (not tested)

samtools view  in.bam | cut -f 12- | tr "\t" "\n"  | grep  "^CR:Z:"  | cut -d ':' -f 3 | sort | uniq | while read S; do samtools view -h in.bam |  awk -v tag="CR:Z:$S" '($0 ~ /^@/ || index($0,tag)>0)' > ${S}.sam ; done

[WouterDeCoster]

Shouldn't this be cut -f 12- rather than cut -f 12?

[Pierre Lindenbaum]

yes, you're right ! thanks, I will fix it

[dccroucher]

Thanks so much for your quick response Pierre! Would you be able to walk me through your loop command, I'm not sure I understand exactly what it's doing...

Thanks :)

[Pierre Lindenbaum]

• samtools view show the bam lines
• cut after column 12 after the QUAL
• tr convert tab to newline
• grep : keep lines starting with CR:Z (oppps there was an error , option -F was not needed)
• cut after 3rd colon to get the value
• sort / uniq to et the distinct tags

for each tag

• print sam with header
• use awk to keep the header and the SAM line containing the tag