Pacbio plasmid contigs does not match with plasmid size through PFGE

Question

[asim]

Hye, I sequenced whole genome of bacteria through pacbio 10kb. I assembled through CANU and got three contigs, two for plasmid and 1 for chromosome. I confirmed this through PlasFlow as well. But the size of my plasmids that i determined through PFGE is 30 kb and 25 kb respectively while the plasmid contigs through CANU are of 41kb and 17kb. Can anyone please guide me how i should proceed (sequencing data) to characterize my plasmids so that my plasmid size and their sequencing matches properly. Thank you!

[GenoMax]

PFGE sizes is likely not final truth (unless you have done multiple gels). That said if you are sure about the sizes then it is possible that the assembly may have moved a portion of the DNA from one plasmid to other. Were the plasmids prepped and sequenced separately?

If you want to be old-fashioned about this then you could do some restriction digests (real/virtual) and see if the fragments you get match from assembled plasmid and real gels.

[asim]

Really appreciate your suggestion, i will repeat PFGE, but should i further polish my assembled data. I mean may be assembled data having repeats.

Answer 1

[Mensur Dlakic]

genomax already gave you good pointers, and I will elaborate on them.

It seems that the sum of plasmid sizes is similar in both cases. PFGE sum is 55 Kb and given the uncertainty of size determination in that range, that's close enough to 58 Kb you get from sequencing. My guess is that the portion of your smaller sequenced plasmid around the area similar to larger plasmid got "transposed" during assembly.

I think restriction digest is the easiest way to solve this. In your sequenced plasmids, find a single restriction enzyme that cuts your plasmids 3-8 times, or a combination of two enzymes that does the same. Then do the actual restriction digest and see what kind of profile you get. Any time you get a match between expected and observed fragment sizes you are confirming the existence of a continuous piece of real plasmid DNA that matches your sequencing result. Now, repeat this exercise as many times as needed with other enzyme combinations and try to figure out what is happening.

Assuming that your observed difference is due to a continuous 8-10 Kb "transposition" from smaller to larger plasmid, a major hint will be if you get a fragment on the gel that is 8-10 Kb smaller than virtual digest predictions. Of course, this can be more complicated than a single-fragment transposition, so you may need to do several independent restriction digests.