I am reading the Whole Genome Sequencing for de novo Assembly Best Practices

  • Use the Sequel II or IIe System and SMRT® Cell 8M to sequence to desired coverage depth for complexity of genome
  • 10- to 15-fold coverage per haplotype recommended

First of all Wikipedia shows different definitions for the term haplotype. Which one applies here?

Second while the recommendations state to use 10- to 15-fold coverage per haplotype I find myself confused about how I would know the haplotype beforehand.

From what I understand of haplotype-estimation I would need the sequenced data first.

How would I go about this when wanting to do a de novo assembly of a species that has no published assembly of any close relatives? That means the only information I would have at that point is the C-value of the species to be sequenced.

Or better, how does the Sequel II figure out the number of haplotypes for achieving the desired fold-coverage?

I hope I'm in the right SE.


They just want to know the ploidy of the sample. In a diploid sample, e.g. humans, there are two haplotypes. This is mostly important for haplotype resolution (i.e. get the two full genomes in a diploid).

Note that elsewhere in the brochure that you link it writes:

Fast and efficient assembly of even the largest genomes with haplotype resolution of complex polyploids

If you are working with a diploid species, that translates to 20-30X coverage. That seems low to me for the older tech of CLR/Sequel sequencing, but they are the experts.

Different recommendations posted here by a service provider suggest 40-50X for structural variant analysis, which seems more reasonable.

  • $\begingroup$ There are still situations where I might not know the ploidity due to a lack of data. At least for now I was not able to find any data on either phylum: Cnidaria or the class: Scyphozoa or the order: Rhizostomeae. Or in the case of Sequoia sempervirens how did they estimate the ploidity in advance? Thanks for the insights! $\endgroup$
    – ilam engl
    Sep 9 at 21:25
  • 1
    $\begingroup$ @ilamengl I think that there are some methods you can look into: 1) karyotype analysis, 2) computational tools, here is one example: bmcbioinformatics.biomedcentral.com/articles/10.1186/… $\endgroup$ Sep 9 at 22:06
  • 1
    $\begingroup$ Additionally, there is a ploidy database which does appear to have cnidarians: cromanpa94.github.io/ACC $\endgroup$ Sep 9 at 22:13

Unless you want to reconstruct all haplotypes, you can sequence to ~30X regardless of ploidy.

  • 1
    $\begingroup$ I would agree for recent autopolyploids, but probably will be less true for polyploids with divergent haplotypes (e.g. allopolyploids). $\endgroup$
    – Kamil S Jaron
    Sep 12 at 13:19
  • $\begingroup$ @KamilSJaron There is not a good way to assemble those genomes for the moment and therefore few know what coverage is needed anyway. 30X would still be a good starting point. $\endgroup$
    – user172818
    Sep 12 at 20:37

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.