Estimating bacterial abundances from SRA files with multiple SRR files

Question

[neel.vadoothker]

If I have an SRA file with multiple SRR files (runs) associated with it, how do I go about analyzing the data? Specifically, can I just use one (the first) SRR file and assume that is all the data I need or do I need to do something like combine/average the SRR files in order to do an analysis?

The analysis I'm trying to perform is estimating bacterial abundances.

Here's a link to the specific dataset I'm looking at: http://www.ncbi.nlm.nih.gov/sra?LinkName=biosample_sra&from_uid=30477

Answer 1

[marina.v.yurieva]

So in your example you are basically trying to get the abundances from 454 16S data. You can simply run qiime 454 pipeline http://qiime.org/tutorials/tutorial.html There is a mothur pipeline as well http://www.mothur.org/wiki/454_SOP

In that experiment they provide V1-V3 and V3-V5 samples, you can pool all the samples of V1-V3s or V3-V5s but you cannot combine these ones together. How much to pool or just to use one sample depends on your needs and your resources.