How To Create A "Discovery Plot"

Question

[Dave Tang]

Hello all,

I just had a question on how to create a "discovery plot", which is a plot to show the detection of genes as a function of sequencing depth.

An example of a discovery plot is figure S5 of this document: http://bloodjournal.hematologylibrary.org/content/suppl/2010/10/19/blood-2010-07-293332.DC1/Document1.pdf

Is it simply a matter of randomly sampling reads, checking which gene this read maps to and tallying to see if this gene was already reported or not?

Thank you in advance,

Dave

[Dave Tang]

To those who may be interested in this, I've written a post on my blog regarding discovery plots and sequencing depth (see http://davetang.org/muse/2013/07/10/how-deep-should-we-sequence/).

Answer 1

[Manu Prestat]

Hi Dave, I'm not sure, but this looks to me very similar to "rarefaction curves" for which you will find a ton of tools.

[Dave Tang]

Hi Manu, ah yes indeed they are similar. I found this tool http://smithlab.usc.edu/plone/software/librarycomplexity after asking the question, and it did exactly what I wanted, in the context of high throughput sequencing. Thank you for your reply and help.