Bacterial genome hybrid assembly

Question

[A_heath]

Hi all,

I have a bacterial genome that I know has an approximate size of 7.3m. I have both short-reads (Illumina) and long-reads (Pacbio). Sequencing showed good coverage. I tried to assemble these data using SPAdes, Flye or Canu but I can not obtain a circular chromosome contig. I know there is plasmid contigs in this genome but the best assembly I got was with Flye (and rounds of Pilon) with a total of 8 contigs.

Do you know another assembler tool that I can use? Do you have any advices on how should I perform this assembly?

Thank you very much for your help,

[andres.firrincieli]

As hybrid assembler you could try Unicycler

[Kumar]

You can try circulator to circularize the assembly.

Answer 1

[oschwengers]

I'd suggest Unicycler. From my experience it sometimes helps to assemble the long reads first using Flye and feed that as a pre-assembly back into Unicycler using both short and long reads. Tricycler would also be an interesting option with various rounds of short read polishing afterwards - but no experience so far with that.

[Kumar]

Can you please elaborate this " the long reads first using Flye and feed that as a pre-assembly back into Unicycler using both short and long reads". Here, you mean long reads is assembly generated by Flye with short reads?

Thank you,

[oschwengers]

I mean, you can use Flye to assemble only the long reads first. Then, the assembly graph of this long-read assembly can be handed over to Unicycler (together with the long and short reads) via

--existing_long_read_assembly

Just try

unicycler --help_all

for a long list of all available options and parameters. For some tricky cases this helped to close the genomes.

[A_heath]

Thank you for your much appreciated help on this! Unicycler worked great, but I still got an un-circularized chromosome. I will try to otpimize this assembly with Tricycler