There should not be a problem as long as you have read counts for each exon. The main issue is that you have little power if the number of reads for a feature is small. So you will need high coverage. You might want to use a normalization method such as the quantile method in edgeR, as I am not sure the others have been tested for this type of data. ( --Naomi At 02:18 PM 3/31/2011, adeonari at mrc-lmb.cam.ac.uk wrote: >Hello Bioconductor community, > >We were wondering if it would be possible to perform differential >expression analysis of exon expression using DESeq or EdgeR. Would the >statistical assumptions be the same, and has anyone attempted this type of >analysis? Any feedback or insights would be really appreciated! > >Cheers, > >Andrew > >_______________________________________________ >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

Hi Vasu, RNA-seq. Cheers, Andrew > You are asking about Affy Exon Expression array or after RNA-seq? > ? > Vasu > > --- On Thu, 3/31/11, adeonari at mrc-lmb.cam.ac.uk > <adeonari at="" mrc-lmb.cam.ac.uk=""> wrote: > > > From: adeonari at mrc-lmb.cam.ac.uk <adeonari at="" mrc-lmb.cam.ac.uk=""> > Subject: [BioC] Using DESeq or EdgeR for Exon Differential Expression > Analysis > To: "'bioconductor at r-project.org'" <bioconductor at="" r-project.org=""> > Date: Thursday, March 31, 2011, 1:18 PM > > > Hello Bioconductor community, > > We were wondering if it would be possible to perform differential > expression analysis of exon expression using DESeq or EdgeR. Would the > statistical assumptions be the same, and has anyone attempted this type of > analysis? Any feedback or insights would be really appreciated! > > Cheers, > > Andrew > > _______________________________________________ > 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 >

There are more than 20% of human exons which have length less than 50 bases. These exons are very likely to have very small number of reads mapped to them. But my understanding is that low count features (such as genes, exons, ...) will be filtered out before the differential expression analysis is performed. Wei On Apr 1, 2011, at 9:59 AM, Naomi Altman wrote: > There should not be a problem as long as you have read counts for each exon. The main issue is that you have little power if the number of reads for a feature is small. So you will need high coverage. You might want to use a normalization method such as the quantile method in edgeR, as I am not sure the others have been tested for this type of data. ( > > --Naomi > > At 02:18 PM 3/31/2011, adeonari at mrc-lmb.cam.ac.uk wrote: >> Hello Bioconductor community, >> >> We were wondering if it would be possible to perform differential >> expression analysis of exon expression using DESeq or EdgeR. Would the >> statistical assumptions be the same, and has anyone attempted this type of >> analysis? Any feedback or insights would be really appreciated! >> >> Cheers, >> >> Andrew >> >> _______________________________________________ >> 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 > > _______________________________________________ > 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}}

Hi Andrew On 03/31/2011 08:18 PM, adeonari at mrc-lmb.cam.ac.uk wrote: > We were wondering if it would be possible to perform differential > expression analysis of exon expression using DESeq or EdgeR. Would the > statistical assumptions be the same, and has anyone attempted this type of > analysis? Any feedback or insights would be really appreciated! As others already remarked: You can simply consider each exon as an individual unit and study it independently from the other exons in the gene. If a gene is differentially expressed, you should see a significant change for all those of its exons that are long enough to have attracted enough counts to achieve significance. However, if you want to see alternative splicing regulation, this might not help to much, as you will need to tease apart the changes in overall gene expression from those that are only affecting the exon under consideration. We are currently working on such a method. Simon