BBSplit syntax for generating builds for the reference genome and how to call different builds.

Question

[aylward.megan]

Can anyone explain how the builds and references should be set in BBSplit?

I have a set of reads that I would like to compare against the target genome and two other contaminant genomes. Can these all be created as one ref directory, each separate or one for target and one for contaminant? If multiple builds are generated should each of these be called in the command line e.g ./bbsplit.sh build=1 build=2 in=reads1.fq in2=reads2.fq ref_x=target.fa ref_y=contam_1.fa,contam2.fa

If anyone has an idea that would be a great help, thanks!

Answer 1

[GenoMax]

See the original thread for that information: http://seqanswers.com/forums/showthread.php?t=41288

For the impatient:

bbsplit.sh in1=reads1.fq in2=reads2.fq ref=ecoli.fa,salmonella.fa basename=out_%.fq outu1=clean1.fq outu2=clean2.fq

Use path= (instead of ref=) if your reference indexes are pre-made.

Output files contain reads that are interleaved by default. They can be separated by

reformat.sh in=out_ecoli.fq out1=out_ecoli_r1.fq out2=out_ecoli_r2.fq

[GiV17]

yes, but after I use BBsplit, I obtain the fastq file and then I can to remap this with STAR but the count? FeatureCount doesn't work well with bbsplit. So my question is: After BBsplit What can I use to map and to calculate the count? Thanks