Shotgun Metagenomics data analysis

[rozitaa]

Hi,

I am new to shotgun metagenomics analysis. I have a project on the human gut microbiota. I know about MG-rast but I want to analyze my data by my own. I have done the quality checking and removal of host contaminated sequences. Now I want to check the taxonomical and functional profile of data. I have seen others use universal single copy marker genes reference. Where can I download these from?
And What is the next step.

Thanks

[fahmida]

Hi,

I am new to shotgun metagenomics analysis. I have a project on the human gut microbiota. I know about MG-rast but I want to analyze my data by my own. I have done the quality checking and removal of host contaminated sequences. Now I want to check the taxonomical and functional profile of data. I have seen others use universal single copy marker genes reference. Where can I download these from?
And What is the next step.

Thanks

Have a look into:

http://qiime.org/

and

http://www.mothur.org/

Reference to databases:

http://qiime.org/home_static/dataFiles.html

This should get you started.

[rozitaa]

Have a look into:

http://qiime.org/

and

http://www.mothur.org/

Reference to databases:

http://qiime.org/home_static/dataFiles.html

This should get you started.

Hi Thanks for quick answer. I have worked with Qiime before. But I mean shotgun sequencing not 16s rRNA analysis!

[sphil]

you can use QIIME (even though it is on your own risk, as they state)

http://qiime.org/tutorials/shotgun_analysis.html

[cliffbeall]

MEGAN might be one way to go, I haven't used it recently but the author seems to actively update it.

There is a different approach here, too:
Kraal, L., et al. (2014). "The prevalence of species and strains in the human microbiome: a resource for experimental efforts." PLoS ONE 9(5): e97279.

Is the marker gene approach you are mentioning MetaPhlan? I don't know much about that.

[bastianwur]

Cross-assembly, then taxonomic annotation (MEGAN or alike), gene calling, annotation, that should be the savest and easiest way to go.