Defining a transcriptome

[madav]

Hi,

I'm working with CLC to map Illumina paired end data from various plant varieties against either a CDS set or an annotated genome, depending on the species, and so to determine tissue transcript expression levels.

Clearly a few mapped reads cannot constitute true transcript expression, so my question is whether there is a generally accepted cutoff filter, based on RPKM or something else that can be used to define the transcriptome for that sample?

thanks in advance

[madav]

hmmm.., not a lot of interest in answering this I see..

However tI hve just found an article that could help address this;

RNA-seq: technical variability and sampling
Lauren M McIntyre, Kenneth K Lopiano, Alison M Morse, Victor Amin, Ann L Oberg, Linda J Young and Sergey V Nuzhdin
BMC Genomics 2011, 12:293 (6 June 2011)