RNA-seq: how many million reads are required ?

[skm]

Hi all,

I am having difficulty in assessing how many million reads are required to optimally capture a transcriptome ?.

I have typically 13 million reads from a Miseq run for 7 samples (1.85 million paired end reads/sample ). Is this number sufficient for carrying out typical RNA-seq analysis using tophat/cufflinks pipeline ? If not what is the optimal number of reads required ?

Thanks!

Regards,
SKM

[mknut]

Depends on species. What organism are the reads from?

[skm]

The organism is a plant (Cajanus cajan).
Any clues ?

Thanks!
SKM

[GenoMax]

Are you only interested in the complete transcriptome or are you trying to answer a specific question (e.g. comparison of test/control)? Is there a related plant species that has been sequenced?

In general it is difficult to estimate the transcriptome size of a species. See post #2: http://seqanswers.com/forums/showthread.php?t=18018

In case of plants ploidy considerations would be another thing to keep in mind. http://gbe.oxfordjournals.org/content/2/534.full

[DonDolowy]

In our lab we are using RNA-seq as an alternative method to microarray. Thus, we are only interested in DEG. In our case we do 8-plexing on mouse samples on a HiSeq, which gives us approximately 20-25mio reads. This is in agreement with e.g. the guidelines for IHEC.