Flye is regarded one of the best assemblers for prokaryotic genomes with high accuracy and least miss-assemblies. We will deal with Flye for now.

mamba activate asm
mamba install flye

flye --nano-hq reads/sample01.fastq -o flye_sample01 -t 10

Then it will magically run on its own. After its completion, check flye_sample01/assembly_info.txt. It may show something like this:

#seq_name	length	cov.	circ.	repeat	mult.	alt_group	graph_path
contig_1	2339103	113	Y	N	1	*	1

The fourth column denotes circ, whether the contig has been circularised. Good job, it is a circular genome!

Sometimes you might see a slightly disappointing one:

#seq_name	length	cov.	circ.	repeat	mult.	alt_group	graph_path
contig_3	1435162	226	N	N	1	*	-1,3,1
contig_2	957190	233	N	N	1	*	1,2,-1
contig_1	37062	394	N	Y	1	*	1

The ideal circular genome is broken into 3 contigs that cannot be stitched together. There might be several reasons to this:

• Difficult to assemble region: see contig_1 has a Y in repeat, meaning there might be a region full of unresolvable repeats
• Reads too short: cannot resolve difficult to assemble regions
• Reads too few (low sequencing coverage): cannot find sufficient overlaps between reads to complete the genome
• Heterogeneous sample: if the reads actually come from multiple strains that have significant genetic difference, assembler would not find a complete path to circularise the genome