Should I Join broken pacbio sequences and use them to scaffold Contigs ?

Question

[elzedleeu]

Hi folks,

I received pacbio sequences produced by sequel system and wish to use the sequences for scaffolding Contigs derived from assembling Illumina data.

I found sequence IDs in the subreads file, which were extracted from subreads.bam file are in the format like this:

"

m54196_171108_070652/6357972/7419_7633 .... ....

m54196_171108_070652/6357972/14702_18160

m54196_171108_070652/6357972/60591_64120

m54196_171108_070652/6357972/64172_66716

"

I guess these sequences are fragments of the same molecule sequenced in the same single ZMW hole. Is this correct? If this is a right guess, should i join them together and used them for genome scaffolding?

Kind Regards,

Elzed

Answer 1

[Istvan Albert]

There is a BAM tag that indicates the ZMW id and thus indicating the same molecule.

I would recommend performing a CSS consensus call - this would merge the reads from the same ZMWs into a higher quality version of them.

Joining (as in concatenating) the sequences would not really make sense in general.

[elzedleeu]

thank you very much Istvan. I agree with you after i read and started to understand pacbio sequencing terminology http://files.pacb.com/software/smrtanalysis/2.2.0/doc/smrtportal/help/!SSL!/Webhelp/Portal_PacBio_Glossary.htm

I will go for ccs even though based on sequence IDs there are not so many molecules got sequenced multiple time.