Question

DESeq variance stabilization

0

Entering edit mode

Lana Schaffer ★ 1.3k

@lana-schaffer-1056

Last seen 10.3 years ago

Hi, I have 2 RNAseq datasets which are the same experiment both having 8 control and 8 treated Samples. I found that they have slightly different standard deviation values and that the Results are different between the 2 datasets. The set with the lower standard deviation Gives a longer list of high adj.pvalues. Can you give me any help here? Lana Schaffer Biostatistics, Informatics DNA Array Core Facility 858-784-2263 [[alternative HTML version deleted]]

RNASeq RNASeq • 1.0k views

ADD COMMENT • link updated 12.3 years ago by Steve Lianoglou ★ 13k • written 12.3 years ago by Lana Schaffer ★ 1.3k

score 0 · Answer 1 · 2012-08-26

0

Entering edit mode

Steve Lianoglou ★ 13k

@steve-lianoglou-2771

Last seen 22 months ago

United States

Hi, On Sun, Aug 26, 2012 at 7:24 AM, Lana Schaffer <schaffer at="" scripps.edu=""> wrote: > Hi, > I have 2 RNAseq datasets which are the same experiment both having 8 control and 8 treated > Samples. I found that they have slightly different standard deviation values and that the > Results are different between the 2 datasets. The set with the lower standard deviation > Gives a longer list of high adj.pvalues. Can you give me any help here? You could try to include the batch as a covariate in your design and run all 16 controls vs 16 treated samples in one go. Is that the type of help you are looking for, or? HTH, -steve -- Steve Lianoglou Graduate Student: Computational Systems Biology | Memorial Sloan-Kettering Cancer Center | Weill Medical College of Cornell University Contact Info: http://cbio.mskcc.org/~lianos/contact

ADD COMMENT • link 12.3 years ago Steve Lianoglou ★ 13k

0

Entering edit mode

Steve, Thank you for your suggestion. However, the idea of these datasets Is to access the reproducability of the results. Lana -----Original Message----- From: Steve Lianoglou [mailto:mailinglist.honeypot@gmail.com] Sent: Sunday, August 26, 2012 5:18 AM To: Lana Schaffer Cc: bioconductor at r-project.org Subject: Re: [BioC] DESeq variance stabilization Hi, On Sun, Aug 26, 2012 at 7:24 AM, Lana Schaffer <schaffer at="" scripps.edu=""> wrote: > Hi, > I have 2 RNAseq datasets which are the same experiment both having 8 > control and 8 treated Samples. I found that they have slightly > different standard deviation values and that the Results are different between the 2 datasets. The set with the lower standard deviation > Gives a longer list of high adj.pvalues. Can you give me any help here? You could try to include the batch as a covariate in your design and run all 16 controls vs 16 treated samples in one go. Is that the type of help you are looking for, or? HTH, -steve -- Steve Lianoglou Graduate Student: Computational Systems Biology | Memorial Sloan- Kettering Cancer Center | Weill Medical College of Cornell University Contact Info: http://cbio.mskcc.org/~lianos/contact

ADD REPLY • link 12.3 years ago Lana Schaffer ★ 1.3k

0

Entering edit mode

Howdy, On Sun, Aug 26, 2012 at 8:19 AM, Lana Schaffer <schaffer at="" scripps.edu=""> wrote: > Steve, > Thank you for your suggestion. However, the idea of these datasets > Is to access the reproducability of the results. So ... are they reproducible? :-) You only mentioned that one of your datasets has "lower standard deviation" and a longer set of differentially expressed genes -- but I guess that would stand to reason, no? Is your assessment of reproducibility strictly a function of the sets of genes called differentially expressed in the two scenarios? If so, is the overlap of the called genes between the two sets particularly poor? How about looking at some type of rank correlation between the two sets of "up" and "down" genes? Perhaps you can look at the top 10, 20, 50, etc. sets of genes and compare like so. -- Steve Lianoglou Graduate Student: Computational Systems Biology | Memorial Sloan-Kettering Cancer Center | Weill Medical College of Cornell University Contact Info: http://cbio.mskcc.org/~lianos/contact

ADD REPLY • link 12.3 years ago Steve Lianoglou ★ 13k