A sample size of 10000 is usually sufficient, unless your data are particularly unusual. For maximum stability and accuracy of prior distribution estimation, you can use 24000 (i.e., every data point) - this will take slightly longer to run but may give very slight improvements. Best wishes, Tom On 08/10/12 11:00, bioconductor-request at r-project.org wrote: > ------------------------------ > > Message: 8 > Date: Mon, 08 Oct 2012 09:50:48 +0000 > From: Fatemehsadat Seyednasrollah<fatsey at="" utu.fi=""> > To:"bioconductor at r-project.org" <bioconductor at="" r-project.org=""> > Subject: [BioC] baySeq and samplesize factor > Message-ID: > <26055A619290434EA444B0C7E5DEFAF4538EFB3C at exch- mbx-01.utu.fi> > Content-Type: text/plain; charset=us-ascii > > Hi all, > > I need to do some RNA seq analysis in order to get DE genes. I want to use baySeq from Bioconductor but as the estimation should be very accurate I wanted to know what is the best "samplesize" amount for function "getPriors.NB" if my dataset has 24000 rows. > > Many thanks in advance. > > > ------------------------------ -- Dr. Thomas J. Hardcastle Department of Plant Sciences University of Cambridge Downing Street Cambridge, CB2 3EA United Kingdom