We will be offering an R workshop December 18-20, 2019. Learn more.

OTU-based approaches

From mothur
Revision as of 18:49, 15 January 2009 by Pschloss (Talk | contribs) (New page: Here we will provide several '''tutorials''' on using the different functions in mothur and example workflows. If you have suggestions for other tutorials, please suggest them in the "dis...)

(diff) ← Older revision | Latest revision (diff) | Newer revision → (diff)
Jump to: navigation, search

Here we will provide several tutorials on using the different functions in mothur and example workflows. If you have suggestions for other tutorials, please suggest them in the "discussion" page. Many of these examples are taken from previously published studies by Schloss and Handelsman using other peoples data. Before launching out on analyzing your own data, we suggest trying these tutorials. mothur is designed to be flexible, so if you can't figure out, how to do something, please let us know!

Identification of OTUs

You have just collected 1,000 (or 100,000) sequences from your favorite environment and you want to know how many OTUs you have sampled. This functionality was previously implemented in DOTUR.

  • Amazon OTUs - How many OTUs did Borneman and Triplett observe after sampling 98 sequences?
  • Pyrosequencing OTUs - How does the number of OTUs change as we change our definition of OTUs?


Single sample analyses

Say you have a 1,000 sequences from a clone library and you are interested in the richness, diversity, evenness of the community or you are interested in determining how much further you need to sample. These are a couple of examples of what you might be interested in doing. This is the type of analysis that was performed by the original DOTUR program and described by Schloss and Handelsman [ref].

  • McCaig richness - Do two Scottish soil samples have the same richness?
  • Schloss richness - Is 1,033 sequences sufficient to obtain a reliable richness estimate?


Multiple sample analyses

Say you obtain 1,000 sequences from each of 10 samples collected from your favorite environment and you are interested in understanding the structure of the community by taking into account all of the samples. Theoretically, you could characterize each of the 10 samples independently, but because you either think that these are replicates or you are interested in how the structure changes as some underlying variable changes you want to perform a composite analysis

  • Human microbiome - Eckburg sampled patients at 7 spots along their GI tract; how did the amount of novel richness change along the GI track?
  • Mouse microbiome - Ley sampled the bacteria found in the cecum of 19 mice. How many bacterial species can be found in mice?


Shared species analyses

Say you have 1,000 sequences from each of 10 samples and you want to generate a Venn diagram or dendrogram to compare their similarity. Here are a couple of tutorials that enable you to generate the data that will enable you to do just that. This is the type of analysis that was performed using both DOTUR and SONS and reported by Schloss and Handelsman.

  • Esophagus comparisons - Pei characterized the microbial communities found in the esophagus of three patients - how much overlap was there?
  • Mouse comparisons - Ley sampled the cecum of related and unrelated mice that were either obese or lean - which mice were most similar?