normalization across sites
1
0
Entering edit mode
Weiwei Shi ★ 1.2k
@weiwei-shi-1407
Last seen 9.7 years ago
Hi, Suppose that I have two sets of arrays, both of which use the same platform. My question is, if I have datasets which have been preprocessed, what's the best way for me to normalize them. Currently I just scale them sample-wise; but not sure of it. Otherwise, should I go back to .CEL file? But how to proceed from there? Thanks, -- Weiwei Shi, Ph.D Research Scientist GeneGO, Inc. "Did you always know?" "No, I did not. But I believed..." ---Matrix III
GO GO • 895 views
ADD COMMENT •
0
Entering edit mode
@sean-davis-490
Last seen 4 months ago
United States
Weiwei Shi wrote: > Hi, > > Suppose that I have two sets of arrays, both of which use the same > platform. My question is, if I have datasets which have been > preprocessed, what's the best way for me to normalize them. Currently > I just scale them sample-wise; but not sure of it. > > Otherwise, should I go back to .CEL file? But how to proceed from there? > Start from raw data, yes. You may need to use an ANOVA or linear model to deal with confounding variables. Quality control is especially important. Sean
ADD COMMENT •
0
Entering edit mode
Ok, if I understand correctly, I should combine two sources of data and run RMA on it? On 6/21/07, Sean Davis <sdavis2 at="" mail.nih.gov=""> wrote: > Weiwei Shi wrote: > > Hi, > > > > Suppose that I have two sets of arrays, both of which use the same > > platform. My question is, if I have datasets which have been > > preprocessed, what's the best way for me to normalize them. Currently > > I just scale them sample-wise; but not sure of it. > > > > Otherwise, should I go back to .CEL file? But how to proceed from there? > > > Start from raw data, yes. You may need to use an ANOVA or linear model > to deal with confounding variables. Quality control is especially > important. > > Sean > -- Weiwei Shi, Ph.D Research Scientist GeneGO, Inc. "Did you always know?" "No, I did not. But I believed..." ---Matrix III
ADD REPLY •
0
Entering edit mode
Yes, you could do that. What Sean is alluding to, is the fact that there is usually a large between-lab variation. Some times this variation is greater than the biological variation. If you are downloading from a website (or for that matter anyway) you should do proper QC of the chips, because you will have little idea of what the original people did with it and whether they made good choices. In principle, if the hybes are good and you use proper normalization, hopefully you will get results that are comparable from lab to lab. But this is not at all given, and you need to be careful when you do these things. Kasper On Jun 21, 2007, at 6:51 PM, Weiwei Shi wrote: > Ok, if I understand correctly, I should combine two sources of data > and run RMA on it? > > On 6/21/07, Sean Davis <sdavis2 at="" mail.nih.gov=""> wrote: >> Weiwei Shi wrote: >>> Hi, >>> >>> Suppose that I have two sets of arrays, both of which use the same >>> platform. My question is, if I have datasets which have been >>> preprocessed, what's the best way for me to normalize them. >>> Currently >>> I just scale them sample-wise; but not sure of it. >>> >>> Otherwise, should I go back to .CEL file? But how to proceed from >>> there? >>> >> Start from raw data, yes. You may need to use an ANOVA or linear >> model >> to deal with confounding variables. Quality control is especially >> important. >> >> Sean >> > > > -- > Weiwei Shi, Ph.D > Research Scientist > GeneGO, Inc. > > "Did you always know?" > "No, I did not. But I believed..." > ---Matrix III > > _______________________________________________ > Bioconductor mailing list > Bioconductor at stat.math.ethz.ch > https://stat.ethz.ch/mailman/listinfo/bioconductor > Search the archives: http://news.gmane.org/ > gmane.science.biology.informatics.conductor
ADD REPLY •
0
Entering edit mode
Then does it mean between-lab variation and between-sample (in one lab) variation could be treated w/o knowing that? I feel there might be more than that. On 6/21/07, Kasper Daniel Hansen <khansen at="" stat.berkeley.edu=""> wrote: > Yes, you could do that. > > What Sean is alluding to, is the fact that there is usually a large > between-lab variation. Some times this variation is greater than the > biological variation. If you are downloading from a website (or for > that matter anyway) you should do proper QC of the chips, because you > will have little idea of what the original people did with it and > whether they made good choices. > > In principle, if the hybes are good and you use proper normalization, > hopefully you will get results that are comparable from lab to lab. > But this is not at all given, and you need to be careful when you do > these things. > > Kasper > > > On Jun 21, 2007, at 6:51 PM, Weiwei Shi wrote: > > > Ok, if I understand correctly, I should combine two sources of data > > and run RMA on it? > > > > On 6/21/07, Sean Davis <sdavis2 at="" mail.nih.gov=""> wrote: > >> Weiwei Shi wrote: > >>> Hi, > >>> > >>> Suppose that I have two sets of arrays, both of which use the same > >>> platform. My question is, if I have datasets which have been > >>> preprocessed, what's the best way for me to normalize them. > >>> Currently > >>> I just scale them sample-wise; but not sure of it. > >>> > >>> Otherwise, should I go back to .CEL file? But how to proceed from > >>> there? > >>> > >> Start from raw data, yes. You may need to use an ANOVA or linear > >> model > >> to deal with confounding variables. Quality control is especially > >> important. > >> > >> Sean > >> > > > > > > -- > > Weiwei Shi, Ph.D > > Research Scientist > > GeneGO, Inc. > > > > "Did you always know?" > > "No, I did not. But I believed..." > > ---Matrix III > > > > _______________________________________________ > > Bioconductor mailing list > > Bioconductor at stat.math.ethz.ch > > https://stat.ethz.ch/mailman/listinfo/bioconductor > > Search the archives: http://news.gmane.org/ > > gmane.science.biology.informatics.conductor > > -- Weiwei Shi, Ph.D Research Scientist GeneGO, Inc. "Did you always know?" "No, I did not. But I believed..." ---Matrix III
ADD REPLY •
0
Entering edit mode
An embedded and charset-unspecified text was scrubbed... Name: not available Url: https://stat.ethz.ch/pipermail/bioconductor/attachments/20070622/ 7b85c65e/attachment.pl
ADD REPLY •

Login before adding your answer.

Similar Posts
Loading Similar Posts
Traffic: 434 users visited in the last hour
Content Search
Users
Tags
Badges
Help About
FAQ
Access RSS
API
Stats

Use of this site constitutes acceptance of our User Agreement and Privacy Policy.

Powered by the version 2.3.6