Pulling bacterial and fungal seq apart

[FGponce]

Hi,
I'm an RNAseq guy who has just started a Genome project. Unfortunately its a baptism of fire since the subject is a symbiosis. I have an assembly using Edena (2K contigs) presumably including both fungal and bacterial fragments.

I have a feeling a meta genomics approach to analysis may be useful.

So what next? Genotype/find the nearest relative by 16s? If so what "generic" 16s should I use to pull out my species 16s (I guess may be more than two things in there).

Is there an established strategy to assemble both separately? Should I go back to the reads and try to separate them prior to an assembly?

Any help appreciated.

FGP

[RickBioinf]

Hi FGP,

You want to use an Illumina HiSeq for this? Or something else? If you use the HiSeq I would say go for the V4 region, this is the best region to use if you are going for just one region. And make pseudoreads out of a Paired End 150.

There are some really good metagenome tools to use for this. I'd suggest to look at QIIME, but if you don't want to install the virtual machine they give for this I wouldn't try to install it. I've tried to install it manually myself and it took me far too long to do and in the end I just stopped doing it because it took too much of my time.
If you want to do it another way, I'd say use BWA or CLCbio.

Hope I have given you some answers

Kind regards,

Rick

[FGponce]

Thanks Rick.

Not obvious from my post but I already have HiSeq2000 data. Just a standard TruSeq DNA V2 library prep. Reads are 2x100bp but likely very low read through to make pseudo reads. This is what I have used to make a first pass de novo genome assembly with Edena.

This was part of a lane that included many other things to make it affordable. I agree best would likely be a MiSeq run with 2x250bps and read through but that will have to wait for a grant to come in.

Hoping this will give me enough preliminary data to land one ; )

FGP.

[RickBioinf]

Hi FGP,

Just wanted to let you know there is an update for the Illumina HiSeq that will let you be able to get 150bp reads, might be handy.

Rick