Hi Stefano,

I always provide 2 files to STAMP when using my own metagenomic profiles; a 'profile file' and a 'metadata file'. I think the latter is only optional if your using profiles downloaded from MG-RAST, for example.

My [tab-del] files look like these:

profile file [where 608, 606, 603 and 601 are my sample IDs]:

domain phylum class 608 606 603 601

Eukaryota Ascomycota Schizosaccharomycetes 20 0 24 0

Bacteria Proteobacteria Deltaproteobacteria 7943 7536 9486 9769

Eukaryota Streptophyta Bryopsida 129 109 135 126

Bacteria Bacteroidetes Flavobacteria 9005 9382 3371 2763

metadata file:

Sample Id Group

601 2

603 2

606 1

608 1

Hope that helps,

Mark

Curse cut-and-paste! Here's what the profile file should look like

Hi Stefano,

this is just a quick answer. I'm no expert in using STAMP nor have i used STAMP for anything but initial testing - but something looks strange in your sample file and from what i can say it seems you're mixing the metagenomic profile and meta data. The profile needs to have one column per sample for each taxonomic category, i.e. six columns with different headers in your example (sample1, ..., sample6). If you want to assign the samples to different groups (like different treatment, conditions or being replicates) you need a second meta data file. This meta data file consist of one column per group (in your case 3 columns, one for the sampleID, one treatment and one replicate) and for each sample a row with the appropriate values. This is essentially the same as other similar software is defining samples and groups (e.g. qiime).

I'm also pretty sure the UserGuide or FAQ covers this topic:

http://kiwi.cs.dal.ca/Software/STAMP

I hope this helps.

@Mark Alston and @Sebastian Jünemann. Thank you for your help! I have loaded metadata file separately and now it works.