Hello Xianfan.

Certainly not an expert on this, but I'll try to help you with what I have figured out. If something is wrong, I hope someone corrects me.

It is not entirely clear from your question what you are trying to achieve. PacBio raw reads (a.k.a polymerase reads) is the continuous stretch sequenced and contains the SMRTbell adapters. Each ZMW produces one polymerase read.

If you want to split the polymerase reads into subreads, you can use the PFilter module from smrtanalysis. Otherwise, if you provide raw polymerase reads to BLASR, it will by default split them into subreads, which you are able to see from the readID in the output: <movie>/<ZMW>/X_Y

• <movie> should not concern you right now, and should be the same for all reads
• <ZMW> corresponds to the well on the SMRTcell where this specific molecule was sequenced
• X, Y correspond to the coordinates on the polymerase read that define the start and end of the subread

Now, concerning the -clipping option in BLASR. This defines how is BLASR gonna report parts of a subread that do not map continuously on the reference. For more information on clipping, please refer to the SAM specification.

dextract of the DEXTRACTOR packages very efficiently converts .bas.h5 to fasta/q and in the process splits reads into subreads.

blasr splits reads into subreads by default, unless you explicitly specify -noSplitSubreads.

And as mpauper mentioned -clipping controls SAM output and not the subread splitting behaviour.