How to estimate a proper depth from the biom summary table?

[Ali shojaee Bakhtiari]

In the following link the authors somehow decided on 114 to be the ideal deapth

http://nbviewer.jupyter.org/github/biocore/qiime/blob/1.9.0/examples/ipynb/illumina_overview_tutorial.ipynb

and claiming that it was deduced from the summary table. Now I have a summary table of my own. How am I supposed to guess the proper depth from it? mean minus one std perhaps? Is there a stnadard process?

Thanks

[Colin Brislawn]

Good evening,

First, a small correction: the authors chose 1114 to be a good depth. 114 reads per sample (10x less) would probably too little in most situations.

Before they choose 1114 to be the cutoff, the authors ran the script 'biom summarize-table'. Have you ran this script yet? It shows you how many reads are in each of your samples. After seeing seeing, this you get to choose which samples to keep and which ones you want to throw away. Like the author say, "so this is always a trade-off between the number of sequences you throw away and the number of samples you throw away."

What does your 'biom summarize-table' look like for your data?

Colin Brislawn

[Ali shojaee Bakhtiari]

Thanks Colin for the reply and the correction. 

The following is the stats of my biom summary table:

$$$$$

Num samples: 4788

Num observations: 45383

Total count: 25894245

Table density (fraction of non-zero values): 0.019

Counts/sample summary:

 Min: 1.0

 Max: 87836.0

 Median: 4445.000

 Mean: 5408.155

 Std. dev.: 4605.644

 Sample Metadata Categories: None provided

 Observation Metadata Categories: Consensus Lineage

$$$$

Mean minus one std leads to about 1000. That where I started from.