Filter out genes with small profile variance from gene expression data

0

Entering edit mode

Jerry Cholo ▴ 190

@jerry-cholo-6218

Last seen 9.5 years ago

Hello, The following are the command lines that I have been using to filter out genes with small profile variance from gene expression data by Matlab. ---------------------------------------------------------------------- % filter genes with small profile variance % less than 25 percentile will be removed from the data mask = genevarfilter(Data,'Percentile',25); % Use the mask as an index into the values to remove the filtered data Data = Data (mask,:); % GeneSymbols located in the first column of the data GeneSymbols = GeneSymbols(mask); ---------------------------------------------------------------------- I need to use R package to perform this task. May someone let me know which functions I could use to do all these tasks? Thanks, Jerry [[alternative HTML version deleted]]

• 2.5k views

ADD COMMENT • link updated 10.1 years ago by Martin Morgan 25k • written 10.1 years ago by Jerry Cholo ▴ 190

0

Entering edit mode

Martin Morgan 25k

@martin-morgan-1513

Last seen 15 hours ago

United States

On 03/09/2014 09:00 PM, Jerry Cholo wrote: > > > Hello, > > > > The following are the command lines that I have been using to filter out > genes with small profile variance from gene expression data by Matlab. The genefilter package varFilter function does what you want; base R's apply function also does this type of operation. Martin > > > > ---------------------------------------------------------------------- > > % filter genes with small profile variance > > % less than 25 percentile will be removed from the data > > mask = genevarfilter(Data,'Percentile',25); > > % Use the mask as an index into the values to remove the filtered data > > Data = Data (mask,:); > > % GeneSymbols located in the first column of the data > > GeneSymbols = GeneSymbols(mask); > > ---------------------------------------------------------------------- > > I need to use R package to perform this task. May someone let me know which > functions I could use to do all these tasks? > > > > Thanks, > > Jerry > > [[alternative HTML version deleted]] > > _______________________________________________ > 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 > -- Computational Biology / Fred Hutchinson Cancer Research Center 1100 Fairview Ave. N. PO Box 19024 Seattle, WA 98109 Location: Arnold Building M1 B861 Phone: (206) 667-2793

ADD COMMENT • link 10.1 years ago Martin Morgan 25k

0

Entering edit mode

Hi! I have been asked to make a candidate list of genes for downstream analysis. We are going to look for secreted proteins into plasma from muscle tissue. I started by picking out genes coding for secretory proteins. Then I sorted by fold change and prediction score, but then I started wondering what actually gives a huge fold change, and now Im not so sure if this is the best approach? Do anyone here have experience with this and have some good ideas? Maybe run GSEA, cluster genes with same expression profile etc? The reason I ask is to try and save money spent on ELISA kits. Thank you very much! Best regards, Sindre

ADD REPLY • link 10.1 years ago Sindre ▴ 110

0

Entering edit mode

Hi, Comments in line: On Mon, Mar 10, 2014 at 11:07 AM, Sindre Lee <sindre.lee at="" studmed.uio.no=""> wrote: > Hi! > I have been asked to make a candidate list of genes for downstream analysis. > We are going to look for secreted proteins into plasma from muscle tissue. I > started by picking out genes coding for secretory proteins. Good start. > Then I sorted by fold change Fold change between what types of samples? > and prediction score What's a prediction score? >, but then I started > wondering what actually gives a huge fold change, What did you come up with? > and now Im not so sure if > this is the best approach? What concerns do you have? > Do anyone here have experience with this and have > some good ideas? Maybe run GSEA, Well, you've already defined a list of secreted proteins relevant to your sample, you can use that gene set and test for enrichment using, eg. camera or roast. > cluster genes with same expression profile What patterns would be interesting to you? > etc? yes, perhaps. > The reason I ask is to try and save money spent on ELISA kits. I think you need to provide more detail regarding what you're trying to study and what data you have to do that with. Is purifying the plasma and running MS on it to identify proteins of interest not a viable option? Is it too expensive? How much is your time worth? -- Steve Lianoglou Computational Biologist Genentech

ADD REPLY • link 10.1 years ago Steve Lianoglou ★ 13k

Login before adding your answer.