Scaffolding for metagenomics

Question

[David]

Hi,

I have a quick question.

I have found a paper describing the different assembly tools http://www.ncbi.nlm.nih.gov/pubmed/24581555 "A comprehensive evaluation of assembly scaffolding tools"

It turns out that they have test abyss-map v1.3.6 which turns out to be a bit old now.

My question is if you recommend to use abyss for metagenomics scaffolding. Right now I'm using megahit and soapdenovo , soapdenovo is not maintained any more so I'm looking for other open source options.

Do you recommend abyss for metagenomics (human microbioe, soil, water...).

Thanks for sharing your experience

david

[benv]

Hi David,

It is probably best to use an assembler/scaffolder that has been specifically designed and tested for metagenomic data. (To the best of my knowledge, ABySS is not well tested for that application.) I did a Google search for "metagenome assembler" and came across the following tools list from https://bioinformatictools.wordpress.com/tag/metagenomic-assembly/

• MetaVelvet
• MetAMOS
• IDBA-UD
• Meta-IDBA
• Ray Meta
• MAP
• Genovo
• Extended Genovo
• SmashCommmunity
• Bambus2
• MetaCAA

Maybe one of those tools will have a scaffolding module that you can use.

Answer 1

[Antonio R. Franco]

I would definitely recommend the using of a bunch of true mate paired sequences under these circumstances to do scaffolding. There are many reasons to rely in their use more than in pure assembly