0
$\begingroup$

I have been trying to assemble my genome via canu. Canu gave me about 200 contigs using default parameters. My genomesize is ~9Mbp. Can you recommend some parameters I can tweak to decrease the contigs number?

$\endgroup$
1
  • $\begingroup$ You can also try other de novo assembly tools like SMARTdenovo to see if they output results more in line with what you expect $\endgroup$
    – Thymine
    Mar 6 '20 at 9:58
1
$\begingroup$

The canu parameter reference guide is always a good place to start.

It would help to know a little bit more about what your assembly looks like. E.g.

  • what is your n50, n90...?
  • What is your dataset coverage?
  • Have you run any tools like quast to evaluate quality?
  • What is your input data type (probably PacBio)? Read length distribution?
  • What is your expected chromosome/replicon number?
  • Have you compared the assembly to any other data, e.g. genomes of close relatives? (quast can do this)

Basically, why do you think that 200 contigs is too many contigs? Sometimes you just don't have enough coverage to get more contiguous/"nice" assemblies. That's not to say it's not possible, of course, just that we don't have enough information to evaluate.

It's possible that some of those contigs are small and artefactual, so you might get rid of them with just length/coverage filtering. You can do some of this in canu itself (see "Output Filtering" section).

$\endgroup$
3
  • $\begingroup$ It is pacbio reads. It results in a contig which is circular and ~9.5 Mbp. Other contigs are either 2Mbp or lesser. They do not seem to be plasmids as well from annotations, and their coverage is also good, ~20X (for the 9.5Mbp contig, its ~30X). Also, other contigs are not repeats of the 9.5 Mbp contig. The longest contig is similar to the reference genome from mauve alignment and dot plot. I am confused if even the 2Mbp contigs could be artefactual? $\endgroup$
    – menten
    Mar 5 '20 at 0:15
  • $\begingroup$ @menten well, to me it sounds like your circular contig is your reference genome, congratulations. the fact that you get other contigs in addition to it suggests to me 1) you have contamination. 2) the other contigs are haplotigs (contigs from alternate haplotypes or clones in your culture) of your genome. I would suggest doing an all-by-all alignment of your contigs via minimap2, which should identify any haplotigs. do you know the ploidy? if it's diploid and not homozygous, well, then you likely have haplotigs with high-coverage pacbio. $\endgroup$ Mar 5 '20 at 19:40
  • $\begingroup$ (i am guessing that you are working on bacteria in which case I know high ploidy may be an odd question, but you never know) $\endgroup$ Mar 5 '20 at 19:41

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.