Triplicate printing and limma
3
0
Entering edit mode
@brooke-powell-elizabeth-1185
Last seen 7.4 years ago
An embedded and charset-unspecified text was scrubbed... Name: not available Url: https://stat.ethz.ch/pipermail/bioconductor/attachments/20050909/ c3559e6f/attachment.pl
• 743 views
0
Entering edit mode
@barry-henderson-49
Last seen 7.4 years ago
Not an expert but I have done this for a custom array with duplicates. Limma can deal with replicate spots that are evenly spaced (ndups and spacing functions in the correlation. Either all of your duplicates are adjascent or spaced by some regular number of interveining spots. I simply duplicate our array layout. The top 2 meta rows are identical to the bottom 2 in my case. My cuplicateCorrelation and lmFit lines then look like: correlation <-duplicateCorrelation(limma.nb, design, ndups=2, spacing=2888) treatmentMeans.fit <- lmFit(limma.nb, design, ndups=2, spacing=2888, correlation=correlation$consensus.correlation) The limitation of limma at this point is that you can not (or I have not figured out how to) deal with the intra array duplication at the same time you want to deal with dye swap. Or at least that is my take on it as a non statistician. Barry -----Original Message----- From: bioconductor-bounces at stat.math.ethz.ch on behalf of Brooke-Powell, Elizabeth Sent: Fri 9/9/2005 11:27 AM To: BlueFutures at yahoogroups.com; bioconductor at stat.math.ethz.ch Cc: Subject: [BioC] Triplicate printing and limma Good Morning, I am writing to ask some advice. We are presently printing a 39,000 feature array and I need some help. We are printing with the triplicates in the same block and I cannot figure out how to get them out of the same block. It is making limma analysis very difficult as the program seems to be designed to analyze printed replicates in different blocks. Can anyone help me in figuring out how to arrange my 96-well plates of oligos and or a get around for limma? Thank you for your help, Liz Brooke-Powell Washington University Medical School, Department of Molecular Microbiology, CB#8230, 660 S. Euclid Ave St. Louis, MO, 63110 [[alternative HTML version deleted]] _______________________________________________ Bioconductor mailing list Bioconductor at stat.math.ethz.ch https://stat.ethz.ch/mailman/listinfo/bioconductor ADD COMMENT 0 Entering edit mode Hi Barry, Thank you for your help.. I have used this for duplicates, but triplicates pose a bit more of a problem. Mainly in the organization of them to get even spacing and different pins. Now maybe limma only really deals with duplicates, that I don't know. The problem is that while my triplicates are space 240 features apart I get an error in LimmaGUI. Now I don't know if by using the command line this will go away or not. I am going to a workshop in a few weeks and hopefully I can learn command line. It just makes me nervous is if I re-sort the data files to have the replicates one after another and do 3 spacing 1 that if the duplicate correlation calculation is actually a modified calculation that accounts for spatial positioning I might be actually performing the wrong calculation. I am not sure if that makes sense? Liz -----Original Message----- From: Barry Henderson [mailto:barry.henderson@ribonomics.com] Sent: Friday, September 09, 2005 10:43 AM To: Brooke-Powell, Elizabeth; BlueFutures at yahoogroups.com; bioconductor at stat.math.ethz.ch Subject: RE: [BioC] Triplicate printing and limma Not an expert but I have done this for a custom array with duplicates. Limma can deal with replicate spots that are evenly spaced (ndups and spacing functions in the correlation. Either all of your duplicates are adjascent or spaced by some regular number of interveining spots. I simply duplicate our array layout. The top 2 meta rows are identical to the bottom 2 in my case. My cuplicateCorrelation and lmFit lines then look like: correlation <-duplicateCorrelation(limma.nb, design, ndups=2, spacing=2888) treatmentMeans.fit <- lmFit(limma.nb, design, ndups=2, spacing=2888, correlation=correlation$consensus.correlation) The limitation of limma at this point is that you can not (or I have not figured out how to) deal with the intra array duplication at the same time you want to deal with dye swap. Or at least that is my take on it as a non statistician. Barry -----Original Message----- From: bioconductor-bounces at stat.math.ethz.ch on behalf of Brooke-Powell, Elizabeth Sent: Fri 9/9/2005 11:27 AM To: BlueFutures at yahoogroups.com; bioconductor at stat.math.ethz.ch Cc: Subject: [BioC] Triplicate printing and limma Good Morning, I am writing to ask some advice. We are presently printing a 39,000 feature array and I need some help. We are printing with the triplicates in the same block and I cannot figure out how to get them out of the same block. It is making limma analysis very difficult as the program seems to be designed to analyze printed replicates in different blocks. Can anyone help me in figuring out how to arrange my 96-well plates of oligos and or a get around for limma? Thank you for your help, Liz Brooke-Powell Washington University Medical School, Department of Molecular Microbiology, CB#8230, 660 S. Euclid Ave St. Louis, MO, 63110 [[alternative HTML version deleted]] _______________________________________________ Bioconductor mailing list Bioconductor at stat.math.ethz.ch https://stat.ethz.ch/mailman/listinfo/bioconductor
0
Entering edit mode
@oosting-j-path-412
Last seen 7.4 years ago
An embedded and charset-unspecified text was scrubbed... Name: not available Url: https://stat.ethz.ch/pipermail/bioconductor/attachments/20050910/ 83ecd178/attachment.pl
0
Entering edit mode
An embedded and charset-unspecified text was scrubbed... Name: not available Url: https://stat.ethz.ch/pipermail/bioconductor/attachments/20050912/ 0ec795ba/attachment.pl
0
Entering edit mode
@gordon-smyth
Last seen 3 hours ago
WEHI, Melbourne, Australia
Dear Liz, I don't think I understand your question. There has never been a requirement in limma that duplicate spots be in different blocks, and it is the same with triplicates. What makes you think there is? The only requirement is that the replicate spots be are equally spaced. If your triplicates are equally spaced, it will make no difference what the actual spacing is. You could re-order them to be at spacing 1, and limma would do the same calculation. The only problem occurs when you have replicates close together *and* far apart on the same arrays. An automatic solution to that situation would be hard. Gordon > Date: Fri, 9 Sep 2005 11:24:24 -0500 > From: etbp2 at borcim.wustl.edu (Brooke-Powell, Elizabeth) > Subject: Re: [BioC] Triplicate printing and limma > To: "'Barry Henderson'" <barry.henderson at="" ribonomics.com="">, > <bluefutures at="" yahoogroups.com="">, <bioconductor at="" stat.math.ethz.ch=""> > > Hi Barry, > > Thank you for your help.. I have used this for duplicates, but triplicates > pose a bit more of a problem. Mainly in the organization of them to get even > spacing and different pins. Now maybe limma only really deals with > duplicates, that I don't know. The problem is that while my triplicates are > space 240 features apart I get an error in LimmaGUI. Now I don't know if by > using the command line this will go away or not. I am going to a workshop in > a few weeks and hopefully I can learn command line. It just makes me nervous > is if I re-sort the data files to have the replicates one after another and > do 3 spacing 1 that if the duplicate correlation calculation is actually a > modified calculation that accounts for spatial positioning I might be > actually performing the wrong calculation. I am not sure if that makes > sense? > > Liz > > -----Original Message----- > From: Barry Henderson [mailto:barry.henderson at ribonomics.com] > Sent: Friday, September 09, 2005 10:43 AM > To: Brooke-Powell, Elizabeth; BlueFutures at yahoogroups.com; > bioconductor at stat.math.ethz.ch > Subject: RE: [BioC] Triplicate printing and limma > > Not an expert but I have done this for a custom array with duplicates. > Limma can deal with replicate spots that are evenly spaced (ndups and > spacing functions in the correlation. Either all of your duplicates are > adjascent or spaced by some regular number of interveining spots. I simply > duplicate our array layout. The top 2 meta rows are identical to the bottom > 2 in my case. My cuplicateCorrelation and lmFit lines then look like: > > correlation <-duplicateCorrelation(limma.nb, design, ndups=2, spacing=2888) > treatmentMeans.fit <- lmFit(limma.nb, design, ndups=2, spacing=2888, > correlation=correlation\$consensus.correlation) > > > The limitation of limma at this point is that you can not (or I have not > figured out how to) deal with the intra array duplication at the same time > you want to deal with dye swap. Or at least that is my take on it as a non > statistician. > > Barry > > -----Original Message----- > From: bioconductor-bounces at stat.math.ethz.ch on behalf of > Brooke-Powell, Elizabeth > Sent: Fri 9/9/2005 11:27 AM > To: BlueFutures at yahoogroups.com; bioconductor at stat.math.ethz.ch > Cc: > Subject: [BioC] Triplicate printing and limma > > > > Good Morning, > > > > I am writing to ask some advice. We are presently printing a 39,000 > feature > array and I need some help. We are printing with the triplicates in > the same > block and I cannot figure out how to get them out of the same block. > It is > making limma analysis very difficult as the program seems to be > designed to > analyze printed replicates in different blocks. Can anyone help me > in > figuring out how to arrange my 96-well plates of oligos and or a get > around > for limma? > > > > Thank you for your help, > > > > Liz Brooke-Powell > > > > Washington University Medical School, > > Department of Molecular Microbiology, > > CB#8230, 660 S. Euclid Ave > > St. Louis, MO, 63110