SEQanswers Papers on the sensitivity of mRNA-seq?
 Register FAQ Members List Calendar Search Today's Posts Mark Forums Read

 Similar Threads Thread Thread Starter Forum Replies Last Post schelhorn RNA Sequencing 5 08-27-2013 03:17 PM kwicher SOLiD 1 09-19-2011 04:45 AM Newsbot! Literature Watch 0 05-24-2011 03:00 AM yjhua2110 Literature Watch 1 02-02-2011 09:50 AM link1 Sample Prep / Library Generation 0 08-12-2010 06:58 PM

 05-10-2010, 10:25 AM #1 liux Member   Location: Midwest Join Date: Mar 2009 Posts: 30 Papers on the sensitivity of mRNA-seq? is there an article discussing the sensitivity of mRNA-seq? I am looking the answer for this question: for a given transcriptome, how many reads of x length are needed to reliably discover a rare transcripts (say 1~2 copies / cell)? Thanks! Last edited by liux; 05-10-2010 at 02:11 PM.
 05-17-2010, 05:46 AM #2 Simon Anders Senior Member   Location: Heidelberg, Germany Join Date: Feb 2010 Posts: 994 Hi I don't know of any papers, but is should be possible to calculate this yourself. Let's say a typical cell has N transcript molecules, then the concentration of your rare transcript is roughly 1/N. If your sequencing run produces M reads (typically, M is up to 20 mio), the probability that a given read is your transcript is M/N. The probability that none of the M reads show your sample is (1-M/N)^M, hence, the probability to see it at least once is, 1-((1-M/N)^M). If you say, you want to see it at least, say, k=10 times, you can easily calculate this with the Poisson distribution. Now, how do you know how many transcripts there are in a cell, i.e., what is the value of N? For such questions, the following nice paper and its web site, that collects a lot of such numbers, might be useful: Phillips and Milo, A feeling for numbers in biology, PNAS, Vol. 106, 21465-71 (2009). Finally, as you are looking for rare transcripts, you might also be interested in this new method to reduce the number of common transcripts, that a colleague happened to have shown me just an hour ago: Bogdanov et al., Normalizing cDNA Libraries, Curr Prot Mol Biol, 5.12.1, Apr 2010 Simon
05-17-2010, 06:04 AM   #3
dariober
Senior Member

Location: Cambridge, UK

Join Date: May 2010
Posts: 311

Quote:
 Originally Posted by liux is there an article discussing the sensitivity of mRNA-seq? I am looking the answer for this question: for a given transcriptome, how many reads of x length are needed to reliably discover a rare transcripts (say 1~2 copies / cell)? Thanks!
Hello,
See if Trapnell et al. 2010 (Nature Biotech) helps. Figure 4 shows how many reads you need to recover a transcript expressed at a given RPKM.
Maybe not exactly what you are asking but possibly you can get a feel for it.

Dario