Constructing single sequence genome assembly out of short reads

Question

[analyst]

My objective is to construct single sequence genome assembly.

But I have short reads only without long reads. I got multiple contigs out of assembly.

My question is that can I use these contigs along short reads to fill the gap and construct single sequence?

Thankyou!

[Asaf]

1. What organism?
2. Why do you want a single genome?

If it's a bacterial genome then most likely contigs will end on rRNA sequences since there are several almost identical copies, it might not be a problem depending on your goals.

[analyst]

1. Its bacteria
2. I want to perform variant calling of reads using assembled genome and compare the results with variant calling result of reads aligned through reference genome.

[Asaf]

So why not do a reference-guided assembly?

[analyst]

Which tool or approach should I use to perform reference guided assembly?

Answer 1

[Philipp Bayer]

It's very unlikely. The assembler did the best it could with the short data you have; any gaps that remain cannot be filled using short reads alone. Some options to create a 'single sequence' from your data:

• add long read data (HiFi, ONT)
• add matepaired data (long insert size short reads)
• add HiC data
• scaffold your contigs using an existing assembly, filling up the gaps with Ns (via Ragout etc https://github.com/fenderglass/Ragout )

But yeah, short reads alone will never go past the draft assembly stage.

[analyst]

Right Phillip. In reference to option 4, scaffolding will fill the gaps with Ns not bases is it okay? Don't we need to add bases instead ?

[Philipp Bayer]

Yes Ns are very common in draft genome assemblies.

You'll get them with matepaired reads too: with matepaired data you have 1000 to 5000 bp between R1 and R2, that's enough to scaffold contigs, but the scaffolder will fill the rest with Ns.