Problems of assembly polishing with RS_Resequencing.1 protocol

[dladuswj]

Hi all,

I’m experiencing a difficulty in getting a complete bacterial genome from PacBio sequencing data using the SMRT Analysis protocols.
My usual workflow is as follows:
De novo assembly by HGAP2 ->Circularization by circlator or minimus -> Polishing by RS_Resequencing.1 protocol until no variants are detected.
This workflow has worked nicely. Usually, repeating RS_Resequencing several times resulted in 100.0% of Consensus Concordance and no variants.

But, for a new bacterial genome we sequenced recently, the result of RS_Resequencing is different from usual. During repeating the RS_Resequencing 8 times, variants have been detected continuously at 10-20 positions and Consensus Concordance never reached 100.0%. At some positions, 1-bp insertion occurred repeatedly per each resequencing.


In my opinion, it seems that more repetition of RS_Resequencing would not be helpful.
What should I do to solve this problem?
Does anyone have a similar experience?

Thanks in advance!

[gconcepcion]

Hi dladuswj,

It seems likely that the bacterial genome that you are sequencing contains more/longer homopolymer/repetitive regions than the typical genomes you are sequencing, which could result in further iterations of the Resequencing algorithm "failing to converge" on a 100% consensus.

You should investigate the alignments at the particular areas where the variants are being called to determine what is causing the variant calling to differ from round to round.

[rhall]

It is possible that the data is from a population that is not perfectly clonal. This happens frequently in slow growing organisms, or things that mutate rapidly.