Hi, in AgiMicroRNa we don´t advocate for any particular method of normalization. We have included scale, quantile and RMA and we let the user decide what might best method for them. You may even want to try some other methods that are not currently implemented in the package. Recently we have published “Processing of Agilent microRNA array data” (López-Romero et al. BMC Research Notes 2010, 3:18) and we concluded that using the AFE signal + quantiles or using the Raw mean signal + RMA method (with not background) might be appropriate methods for the preprocessing of the microRNA signals. Although in that paper we focused mainly in the way that the replicated probes are summarized in to a single expression measure and we only assessed the methods in terms of the signal variability. According to Agilent the controls cannot be used for normalization because they don’t behave consistenly, but this is something that I haven´t checked myself. You can try what Wei suggests. p.- On Fri, Feb 12, 2010 at 12:02 AM, Wei Shi <shi@wehi.edu.au> wrote: > Dear Daniel: > > You might need to check the quality of these negative controls to see if > they provide a good measurement of the background noise. One way to do this > is to use the density function to check if these probes are normally > distributed. > > If the negative control probes are a good representation of the > background noise, you can consider using these controls to estimate the > background mean intensity and the standard deviation of the background > intensities which are two of the parameters needed by the normexp background > correction model. The limma function neqc uses the negative controls and > positive controls to do the normalization which might be useful for your > data. > > Cheers, > Wei > > > Daniel Brewer wrote: > >> Hello, >> >> I have been trying to work out the best way to normalise microRNA >> microarrays recently as there seems to be some debate about whether the >> normal mRNA assumptions hold. >> >> The AgiMicroRna suggests using quantile or scale, whereas other posts I >> have read have suggested VSN. >> >> My question is whether the control probes on the Agilent chip could be >> used o do the normalisation i.e. the Negative and Positive Control probes. >> >> Would this be a good idea? And how would one go about this? A loess >> using the weights to define what probes are used? >> >> Thanks >> >> Dan >> >> >> > > ______________________________________________________________________ > The information in this email is confidential and intend...{{dropped:4}} > > > _______________________________________________ > Bioconductor mailing list > Bioconductor@stat.math.ethz.ch > https://stat.ethz.ch/mailman/listinfo/bioconductor > Search the archives: > http://news.gmane.org/gmane.science.biology.informatics.conductor > [[alternative HTML version deleted]]