Hi Henry: An alternative way is to perform a gene-oriented analysis in which you can choose a specific region for each gene to examine the binding enrichment within it. For example, you can choose a region around the transcription start site of gene to look for the enrichment of transcription factor biding sites or even histone marks like H3K4me etc. You may also find it useful to estimate biological variation between your test sample and control sample to remove sites which are found to be enriched due to the biological variation. The edgeR package can help you do this (?estimateGLMCommonDisp). Cheers, Wei On May 11, 2011, at 1:47 AM, Henry Paik wrote: > Hello Alicia and Wei, > > Wei, > > Thanks for your email. It is helpful. > > Alicia, > > Good points. Yes, I thought of using bins. However, since we have the input (control) data, I thought I would be able to clean up the data a little bit before normalization by some kind of filtering/masking. Of course, bin sizes would be important but maybe we can find an optimal bin size? > I had run a peak-calling program on those samples with a input data and got the results. But when I read that the program scaled "linearly" to normalize, I wanted to try a different approach. > > I am still new to ChIP-Seq so please enlighten my ignorance. > > Thanks! > > - Henry > > > On 05/10/2011 07:44 AM, Alicia Oshlack wrote: >> Hi Henry, >> >> In my view it's not obvious how you would perform quantile normalization on ChIP-seq data. Were you planning to bin the data then quantile normalize binned counts? Then you run into the issue of what bin size to use for all you analysis and most peak-finding programs will not take binned data. Even if you use each base as a bin we know there are base specific biases so it's not clear that this would be effective especially as there will be very many 0 and 1 counts. >> >> Cheers, >> Alicia >> >> >> Date: Tue, 10 May 2011 09:13:18 +1000 >> >> Hi Henry: >> >> There are a few BioC packages which can perform quantile normalization. One of them is the normalizeQuantiles function in limma. >> >> Cheers, >> Wei >> >> >> >> On May 10, 2011, at 12:15 AM, Henry Paik wrote: >> >>> Hello BioC, >>> >>> I am wondering if there is a package or example codes to quantile- normalize ChIP-Seq data. >>> >>> I would like to compare Sample vs. Sample and think it would be nice to use quantile normalization - sort of we use it for microarray. >>> >>> I looked up a couple of tools - MACS and QuEST. It seems that they normalize by "scaling linearly." >>> >>> Sequencing and microarray do not share the exact features but for normalization, wouldn't it be okay to use quantile normalization? >>> >>> Thanks! >>> >>> - Henry >>> >> ______________________________________________________________________ >> This email has been scanned by the MessageLabs Email Security System. >> For more information please visit http://www.messagelabs.com/email >> ______________________________________________________________________ > > _______________________________________________ > Bioconductor mailing list > Bioconductor at r-project.org > https://stat.ethz.ch/mailman/listinfo/bioconductor > Search the archives: http://news.gmane.org/gmane.science.biology.informatics.conductor ______________________________________________________________________ The information in this email is confidential and intend...{{dropped:6}}