Limma question_Intra-Spot Correlation Question
1
0
Entering edit mode
@jgomezuni-potsdamde-4586
Last seen 10.2 years ago
Dear List, I have a set of Agilent chips that I would like to analyse following the procedure for separate channel analysis. I just want to add that following the brief procedure outline in Limma User Guide wasn't useful. I have made the targets file as usual FileName Cy3 Cy5 1 US83800208_252412610022_1_4.txt WT_4 OX_4 2 US83800208_252412610019_1_1.txt KD_4 WT_4 3 US83800208_252412610019_1_2.txt OX_4 KD_4 4 US83800208_252412610019_1_3.txt OX_21 OX_4 5 US83800208_252412610019_1_4.txt WT_21 WT_4 6 US83800208_252412610020_1_1.txt KD_4 KD_21 7 US83800208_252412610020_1_2.txt WT_21 OX_21 8 US83800208_252412610021_2_1.txt KD_21 WT_21 9 US83800208_252412610020_1_3.txt OX_21 KD_21 I read the chips using read.maimages and creating an RG object along the way. Then I normalized the arrays using Aquantile (Normalization Between Arrays). I convert the targets file and then looks like this: channel.col FileName Target 1.1 1 US83800208_252412610022_1_4.txt WT_4 1.2 2 US83800208_252412610022_1_4.txt OX_4 2.1 1 US83800208_252412610019_1_1.txt KD_4 2.2 2 US83800208_252412610019_1_1.txt WT_4 3.1 1 US83800208_252412610019_1_2.txt OX_4 3.2 2 US83800208_252412610019_1_2.txt KD_4 4.1 1 US83800208_252412610019_1_3.txt OX_21 4.2 2 US83800208_252412610019_1_3.txt OX_4 5.1 1 US83800208_252412610019_1_4.txt WT_21 5.2 2 US83800208_252412610019_1_4.txt WT_4 6.1 1 US83800208_252412610020_1_1.txt KD_4 6.2 2 US83800208_252412610020_1_1.txt KD_21 7.1 1 US83800208_252412610020_1_2.txt WT_21 7.2 2 US83800208_252412610020_1_2.txt OX_21 8.1 1 US83800208_252412610021_2_1.txt KD_21 8.2 2 US83800208_252412610021_2_1.txt WT_21 9.1 1 US83800208_252412610020_1_3.txt OX_21 9.2 2 US83800208_252412610020_1_3.txt KD_21 When using the function intraspotCorrelation I got an error regarding " Missing or infinite values found in M or A". Checking older post regarding the erro, it may be becaouse few probes have after normalization negative intensities. I would like to know if the process I started is the right one for this kind of analysis and second If there is a kind of filter I can use in limma to get rid of those neg intensities to proceed to the next step. Thanks in advance for any help. Cheers, Judy -- Judith Lucia Gomez, PhD Centre for Plant Biotechnology and Genomics - CBGP 28223 Pozuelo de Alarc?n (Madrid) Spain
Normalization limma PROcess convert Normalization limma PROcess convert • 1.5k views
ADD COMMENT
0
Entering edit mode
@gordon-smyth
Last seen 26 minutes ago
WEHI, Melbourne, Australia
Dear Judith, Getting negative intensities has nothing to do with normalization, but everything to do with background correction. Please see page 24 of the limma User's Guide: "we often find RG <- backgroundCorrect(RG, method="normexp", offset=50) to be preferable to the simple background subtraction when using output from most image analysis programs. This method adjusts the foreground adaptively for the background intensities and results in strictly positive adjusted intensities, i.e., negative or zero corrected intensities are avoided." Best wishes Gordon > Message: 10 > Date: Sun, 05 Feb 2012 11:45:12 +0100 > From: jgomez at uni-potsdam.de > To: Bioconductor mailing list <bioconductor at="" r-project.org=""> > Subject: [BioC] Limma question_Intra-Spot Correlation Question > > Dear List, > I have a set of Agilent chips that I would like to analyse following > the procedure for separate channel analysis. > I just want to add that following the brief procedure outline in Limma > User Guide wasn't useful. > I have made the targets file as usual > FileName Cy3 Cy5 > 1 US83800208_252412610022_1_4.txt WT_4 OX_4 > 2 US83800208_252412610019_1_1.txt KD_4 WT_4 > 3 US83800208_252412610019_1_2.txt OX_4 KD_4 > 4 US83800208_252412610019_1_3.txt OX_21 OX_4 > 5 US83800208_252412610019_1_4.txt WT_21 WT_4 > 6 US83800208_252412610020_1_1.txt KD_4 KD_21 > 7 US83800208_252412610020_1_2.txt WT_21 OX_21 > 8 US83800208_252412610021_2_1.txt KD_21 WT_21 > 9 US83800208_252412610020_1_3.txt OX_21 KD_21 > I read the chips using read.maimages and creating an RG object along > the way. Then I normalized the arrays using Aquantile (Normalization > Between Arrays). I convert the targets file and then looks like this: > channel.col FileName Target > 1.1 1 US83800208_252412610022_1_4.txt WT_4 > 1.2 2 US83800208_252412610022_1_4.txt OX_4 > 2.1 1 US83800208_252412610019_1_1.txt KD_4 > 2.2 2 US83800208_252412610019_1_1.txt WT_4 > 3.1 1 US83800208_252412610019_1_2.txt OX_4 > 3.2 2 US83800208_252412610019_1_2.txt KD_4 > 4.1 1 US83800208_252412610019_1_3.txt OX_21 > 4.2 2 US83800208_252412610019_1_3.txt OX_4 > 5.1 1 US83800208_252412610019_1_4.txt WT_21 > 5.2 2 US83800208_252412610019_1_4.txt WT_4 > 6.1 1 US83800208_252412610020_1_1.txt KD_4 > 6.2 2 US83800208_252412610020_1_1.txt KD_21 > 7.1 1 US83800208_252412610020_1_2.txt WT_21 > 7.2 2 US83800208_252412610020_1_2.txt OX_21 > 8.1 1 US83800208_252412610021_2_1.txt KD_21 > 8.2 2 US83800208_252412610021_2_1.txt WT_21 > 9.1 1 US83800208_252412610020_1_3.txt OX_21 > 9.2 2 US83800208_252412610020_1_3.txt KD_21 > When using the function intraspotCorrelation I got an error regarding > " Missing or infinite values found in M or A". > Checking older post regarding the erro, it may be becaouse few probes > have after normalization negative intensities. > I would like to know if the process I started is the right one for > this kind of analysis and second If there is a kind of filter I can > use in limma to get rid of those neg intensities to proceed to the > next step. > Thanks in advance for any help. > Cheers, > Judy > > -- > Judith Lucia Gomez, PhD > Centre for Plant Biotechnology and Genomics - CBGP > 28223 Pozuelo de Alarc?n (Madrid) > Spain ______________________________________________________________________ The information in this email is confidential and intend...{{dropped:4}}
ADD COMMENT
0
Entering edit mode
Dear Gordon and limma users, I am trying to fit a linear model to assess variability in healthy donors over days for both genders using Affymetrix microarray data. I am interested in assessing variability arising from geneder, day of blood donated and individual donors. When I tried to do a linear fit in using limma, I am not able to assess the coefficients for gender.System is computationally singular for the gender main effect. Is there any way to assess the variability arising from all three main effects? Any help would be greatly appreciated. Following is my target file: Array_NUID gender day donor NUID.0000.0133.2260 male d1 M1 NUID.0000.0133.2261 male d2 M1 NUID.0000.0133.2262 male d8 M1 NUID.0000.0133.2230 female d1 F2 NUID.0000.0133.6809 female d2 F2 NUID.0000.0133.2251 female d8 F2 NUID.0000.0133.2228 female d15 F2 NUID.0000.0133.2229 female d28 F2 NUID.0000.0133.2231 female d1 F3 NUID.0000.0133.2232 female d2 F3 NUID.0000.0133.2233 female d8 F3 NUID.0000.0133.2234 female d15 F3 NUID.0000.0133.2265 male d1 M4 NUID.0000.0133.2267 male d8 M4 NUID.0000.0133.2268 male d15 M4 NUID.0000.0133.2269 male d28 M4 NUID.0000.0133.2236 female d1 F5 NUID.0000.0133.2237 female d2 F5 NUID.0000.0133.2238 female d8 F5 NUID.0000.0133.6810 female d15 F5 NUID.0000.0133.2240 female d28 F5 NUID.0000.0133.2270 male d1 M6 NUID.0000.0133.2271 male d2 M6 NUID.0000.0133.2272 male d8 M6 NUID.0000.0133.2273 male d15 M6 NUID.0000.0133.2274 male d28 M6 NUID.0000.0133.2241 female d1 F7 NUID.0000.0133.2242 female d2 F7 NUID.0000.0133.2243 female d8 F7 NUID.0000.0133.2244 female d15 F7 NUID.0000.0133.2245 female d28 F7 NUID.0000.0133.2246 female d1 F8 NUID.0000.0133.2247 female d2 F8 NUID.0000.0133.2248 female d8 F8 NUID.0000.0133.2249 female d15 F8 NUID.0000.0133.2250 female d28 F8 Following is the code I am using: library("limma") rna <- read.table('qced_non_redundant_rma_data.txt', header=T, row.names=1, sep= "\t") annotation <- read.table('Test_Samples_Annotation.txt', header=T, row.names=1, sep= "\t") ## rearrange the annotations to conform to the RMA data annotation <- annotation[colnames(rna),] ## check that the annotations and data match if (sum(sort(rownames(annotation)) == sort(colnames(rna))) != length(rownames(annotation))) { stop ("ERROR: annotation rownames and RMA colnames do not match") } donor <- factor(annotation$donor) day <- factor(annotation$day) gender <- factor(annotation$gender) design <- model.matrix(~ donor + gender + day, data=annotation) fit <- lmFit(rna,design) fit2 <- eBayes(fit) write.fit(fit,file="effects.txt",digits=30, adjust="BH", method="separate",sep="\t") Following is my sessionInfo: > sessionInfo() R version 2.10.1 (2009-12-14) x86_64-unknown-linux-gnu locale: [1] LC_CTYPE=en_US.iso885915 LC_NUMERIC=C [3] LC_TIME=en_US.iso885915 LC_COLLATE=en_US.iso885915 [5] LC_MONETARY=C LC_MESSAGES=en_US.iso885915 [7] LC_PAPER=en_US.iso885915 LC_NAME=C [9] LC_ADDRESS=C LC_TELEPHONE=C [11] LC_MEASUREMENT=en_US.iso885915 LC_IDENTIFICATION=C attached base packages: [1] stats graphics grDevices utils datasets methods base other attached packages: [1] limma_3.2.3 loaded via a namespace (and not attached): [1] tools_2.10.1 [[alternative HTML version deleted]]
ADD REPLY
0
Entering edit mode
The problem is that donor is nested within gender. The only way I am aware of to handle this in limma is to use donor as a block. Regards, Naomi At 11:24 AM 2/7/2012, somnath bandyopadhyay wrote: > Dear Gordon and limma users, > >I am trying to fit a linear model to assess variability in healthy >donors over days for both genders using Affymetrix microarray data. >I am interested in assessing variability arising from geneder, day >of blood donated and individual donors. When I tried to do a linear >fit in using limma, I am not able to assess the coefficients for >gender.System is computationally singular for the gender main >effect. Is there any way to assess the variability arising from all >three main effects? Any help would be greatly appreciated. > >Following is my target file: > > > > > > > > >Array_NUID >gender >day >donor > >NUID.0000.0133.2260 >male >d1 >M1 > >NUID.0000.0133.2261 >male >d2 >M1 > >NUID.0000.0133.2262 >male >d8 >M1 > >NUID.0000.0133.2230 >female >d1 >F2 > >NUID.0000.0133.6809 >female >d2 >F2 > >NUID.0000.0133.2251 >female >d8 >F2 > >NUID.0000.0133.2228 >female >d15 >F2 > >NUID.0000.0133.2229 >female >d28 >F2 > >NUID.0000.0133.2231 >female >d1 >F3 > >NUID.0000.0133.2232 >female >d2 >F3 > >NUID.0000.0133.2233 >female >d8 >F3 > >NUID.0000.0133.2234 >female >d15 >F3 > >NUID.0000.0133.2265 >male >d1 >M4 > >NUID.0000.0133.2267 >male >d8 >M4 > >NUID.0000.0133.2268 >male >d15 >M4 > >NUID.0000.0133.2269 >male >d28 >M4 > >NUID.0000.0133.2236 >female >d1 >F5 > >NUID.0000.0133.2237 >female >d2 >F5 > >NUID.0000.0133.2238 >female >d8 >F5 > >NUID.0000.0133.6810 >female >d15 >F5 > >NUID.0000.0133.2240 >female >d28 >F5 > >NUID.0000.0133.2270 >male >d1 >M6 > >NUID.0000.0133.2271 >male >d2 >M6 > >NUID.0000.0133.2272 >male >d8 >M6 > >NUID.0000.0133.2273 >male >d15 >M6 > >NUID.0000.0133.2274 >male >d28 >M6 > >NUID.0000.0133.2241 >female >d1 >F7 > >NUID.0000.0133.2242 >female >d2 >F7 > >NUID.0000.0133.2243 >female >d8 >F7 > >NUID.0000.0133.2244 >female >d15 >F7 > >NUID.0000.0133.2245 >female >d28 >F7 > >NUID.0000.0133.2246 >female >d1 >F8 > >NUID.0000.0133.2247 >female >d2 >F8 > >NUID.0000.0133.2248 >female >d8 >F8 > >NUID.0000.0133.2249 >female >d15 >F8 > >NUID.0000.0133.2250 >female >d28 >F8 > > >Following is the code I am using: >library("limma") >rna <- read.table('qced_non_redundant_rma_data.txt', header=T, >row.names=1, sep= "\t") >annotation <- read.table('Test_Samples_Annotation.txt', header=T, >row.names=1, sep= "\t") >## rearrange the annotations to conform to the RMA data >annotation <- annotation[colnames(rna),] >## check that the annotations and data match >if (sum(sort(rownames(annotation)) == sort(colnames(rna))) != >length(rownames(annotation))) { > stop ("ERROR: annotation rownames and RMA colnames do not match") >} >donor <- factor(annotation$donor) >day <- factor(annotation$day) >gender <- factor(annotation$gender) >design <- model.matrix(~ donor + gender + day, data=annotation) >fit <- lmFit(rna,design) >fit2 <- eBayes(fit) >write.fit(fit,file="effects.txt",digits=30, adjust="BH", >method="separate",sep="\t") > > > >Following is my sessionInfo: > > sessionInfo() >R version 2.10.1 (2009-12-14) >x86_64-unknown-linux-gnu >locale: > [1] LC_CTYPE=en_US.iso885915 LC_NUMERIC=C > [3] LC_TIME=en_US.iso885915 LC_COLLATE=en_US.iso885915 > [5] LC_MONETARY=C LC_MESSAGES=en_US.iso885915 > [7] LC_PAPER=en_US.iso885915 LC_NAME=C > [9] LC_ADDRESS=C LC_TELEPHONE=C >[11] LC_MEASUREMENT=en_US.iso885915 LC_IDENTIFICATION=C >attached base packages: >[1] stats graphics grDevices utils datasets methods base >other attached packages: >[1] limma_3.2.3 >loaded via a namespace (and not attached): >[1] tools_2.10.1 > [[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
ADD REPLY
0
Entering edit mode
Thanks Naomi. > Date: Tue, 7 Feb 2012 20:32:56 -0500 > To: genome1976@hotmail.com; smyth@wehi.edu.au > From: naomi@stat.psu.edu > Subject: Re: [BioC] Coefficients not estimable > CC: bioconductor@r-project.org > > The problem is that donor is nested within gender. The only way I am > aware of to handle this in limma is to use donor as a block. > > Regards, > Naomi > > > At 11:24 AM 2/7/2012, somnath bandyopadhyay wrote: > > > > Dear Gordon and limma users, > > > >I am trying to fit a linear model to assess variability in healthy > >donors over days for both genders using Affymetrix microarray data. > >I am interested in assessing variability arising from geneder, day > >of blood donated and individual donors. When I tried to do a linear > >fit in using limma, I am not able to assess the coefficients for > >gender.System is computationally singular for the gender main > >effect. Is there any way to assess the variability arising from all > >three main effects? Any help would be greatly appreciated. > > > >Following is my target file: > > > > > > > > > > > > > > > > > >Array_NUID > >gender > >day > >donor > > > >NUID.0000.0133.2260 > >male > >d1 > >M1 > > > >NUID.0000.0133.2261 > >male > >d2 > >M1 > > > >NUID.0000.0133.2262 > >male > >d8 > >M1 > > > >NUID.0000.0133.2230 > >female > >d1 > >F2 > > > >NUID.0000.0133.6809 > >female > >d2 > >F2 > > > >NUID.0000.0133.2251 > >female > >d8 > >F2 > > > >NUID.0000.0133.2228 > >female > >d15 > >F2 > > > >NUID.0000.0133.2229 > >female > >d28 > >F2 > > > >NUID.0000.0133.2231 > >female > >d1 > >F3 > > > >NUID.0000.0133.2232 > >female > >d2 > >F3 > > > >NUID.0000.0133.2233 > >female > >d8 > >F3 > > > >NUID.0000.0133.2234 > >female > >d15 > >F3 > > > >NUID.0000.0133.2265 > >male > >d1 > >M4 > > > >NUID.0000.0133.2267 > >male > >d8 > >M4 > > > >NUID.0000.0133.2268 > >male > >d15 > >M4 > > > >NUID.0000.0133.2269 > >male > >d28 > >M4 > > > >NUID.0000.0133.2236 > >female > >d1 > >F5 > > > >NUID.0000.0133.2237 > >female > >d2 > >F5 > > > >NUID.0000.0133.2238 > >female > >d8 > >F5 > > > >NUID.0000.0133.6810 > >female > >d15 > >F5 > > > >NUID.0000.0133.2240 > >female > >d28 > >F5 > > > >NUID.0000.0133.2270 > >male > >d1 > >M6 > > > >NUID.0000.0133.2271 > >male > >d2 > >M6 > > > >NUID.0000.0133.2272 > >male > >d8 > >M6 > > > >NUID.0000.0133.2273 > >male > >d15 > >M6 > > > >NUID.0000.0133.2274 > >male > >d28 > >M6 > > > >NUID.0000.0133.2241 > >female > >d1 > >F7 > > > >NUID.0000.0133.2242 > >female > >d2 > >F7 > > > >NUID.0000.0133.2243 > >female > >d8 > >F7 > > > >NUID.0000.0133.2244 > >female > >d15 > >F7 > > > >NUID.0000.0133.2245 > >female > >d28 > >F7 > > > >NUID.0000.0133.2246 > >female > >d1 > >F8 > > > >NUID.0000.0133.2247 > >female > >d2 > >F8 > > > >NUID.0000.0133.2248 > >female > >d8 > >F8 > > > >NUID.0000.0133.2249 > >female > >d15 > >F8 > > > >NUID.0000.0133.2250 > >female > >d28 > >F8 > > > > > >Following is the code I am using: > >library("limma") > >rna <- read.table('qced_non_redundant_rma_data.txt', header=T, > >row.names=1, sep= "\t") > >annotation <- read.table('Test_Samples_Annotation.txt', header=T, > >row.names=1, sep= "\t") > >## rearrange the annotations to conform to the RMA data > >annotation <- annotation[colnames(rna),] > >## check that the annotations and data match > >if (sum(sort(rownames(annotation)) == sort(colnames(rna))) != > >length(rownames(annotation))) { > > stop ("ERROR: annotation rownames and RMA colnames do not match") > >} > >donor <- factor(annotation$donor) > >day <- factor(annotation$day) > >gender <- factor(annotation$gender) > >design <- model.matrix(~ donor + gender + day, data=annotation) > >fit <- lmFit(rna,design) > >fit2 <- eBayes(fit) > >write.fit(fit,file="effects.txt",digits=30, adjust="BH", > >method="separate",sep="\t") > > > > > > > >Following is my sessionInfo: > > > sessionInfo() > >R version 2.10.1 (2009-12-14) > >x86_64-unknown-linux-gnu > >locale: > > [1] LC_CTYPE=en_US.iso885915 LC_NUMERIC=C > > [3] LC_TIME=en_US.iso885915 LC_COLLATE=en_US.iso885915 > > [5] LC_MONETARY=C LC_MESSAGES=en_US.iso885915 > > [7] LC_PAPER=en_US.iso885915 LC_NAME=C > > [9] LC_ADDRESS=C LC_TELEPHONE=C > >[11] LC_MEASUREMENT=en_US.iso885915 LC_IDENTIFICATION=C > >attached base packages: > >[1] stats graphics grDevices utils datasets methods base > >other attached packages: > >[1] limma_3.2.3 > >loaded via a namespace (and not attached): > >[1] tools_2.10.1 > > [[alternative HTML version deleted]] > > > >_______________________________________________ > >Bioconductor mailing list > >Bioconductor@r-project.org > >https://stat.ethz.ch/mailman/listinfo/bioconductor > >Search the archives: > >http://news.gmane.org/gmane.science.biology.informatics.conductor > > > [[alternative HTML version deleted]]
ADD REPLY
0
Entering edit mode
Or perhaps as a nested fixed model model.matrix(~gender+gender:donor+day) Whether to fit donor as a fixed term or as a block really depends on what questions you want to answer and, in particular, what you want to interpret from a donor "main effect". Best wishes Gordon --------------------------------------------- Professor Gordon K Smyth, Bioinformatics Division, Walter and Eliza Hall Institute of Medical Research, 1G Royal Parade, Parkville, Vic 3052, Australia. smyth at wehi.edu.au http://www.wehi.edu.au http://www.statsci.org/smyth On Tue, 7 Feb 2012, Naomi Altman wrote: > The problem is that donor is nested within gender. The only way I am aware > of to handle this in limma is to use donor as a block. > > Regards, > Naomi > > > At 11:24 AM 2/7/2012, somnath bandyopadhyay wrote: > > >> Dear Gordon and limma users, >> >> I am trying to fit a linear model to assess variability in healthy donors >> over days for both genders using Affymetrix microarray data. I am >> interested in assessing variability arising from geneder, day of blood >> donated and individual donors. When I tried to do a linear fit in using >> limma, I am not able to assess the coefficients for gender.System is >> computationally singular for the gender main effect. Is there any way to >> assess the variability arising from all three main effects? Any help would >> be greatly appreciated. >> >> Following is my target file: >> >> >> >> >> >> >> >> >> Array_NUID >> gender >> day >> donor >> >> NUID.0000.0133.2260 >> male >> d1 >> M1 >> >> NUID.0000.0133.2261 >> male >> d2 >> M1 >> >> NUID.0000.0133.2262 >> male >> d8 >> M1 >> >> NUID.0000.0133.2230 >> female >> d1 >> F2 >> >> NUID.0000.0133.6809 >> female >> d2 >> F2 >> >> NUID.0000.0133.2251 >> female >> d8 >> F2 >> >> NUID.0000.0133.2228 >> female >> d15 >> F2 >> >> NUID.0000.0133.2229 >> female >> d28 >> F2 >> >> NUID.0000.0133.2231 >> female >> d1 >> F3 >> >> NUID.0000.0133.2232 >> female >> d2 >> F3 >> >> NUID.0000.0133.2233 >> female >> d8 >> F3 >> >> NUID.0000.0133.2234 >> female >> d15 >> F3 >> >> NUID.0000.0133.2265 >> male >> d1 >> M4 >> >> NUID.0000.0133.2267 >> male >> d8 >> M4 >> >> NUID.0000.0133.2268 >> male >> d15 >> M4 >> >> NUID.0000.0133.2269 >> male >> d28 >> M4 >> >> NUID.0000.0133.2236 >> female >> d1 >> F5 >> >> NUID.0000.0133.2237 >> female >> d2 >> F5 >> >> NUID.0000.0133.2238 >> female >> d8 >> F5 >> >> NUID.0000.0133.6810 >> female >> d15 >> F5 >> >> NUID.0000.0133.2240 >> female >> d28 >> F5 >> >> NUID.0000.0133.2270 >> male >> d1 >> M6 >> >> NUID.0000.0133.2271 >> male >> d2 >> M6 >> >> NUID.0000.0133.2272 >> male >> d8 >> M6 >> >> NUID.0000.0133.2273 >> male >> d15 >> M6 >> >> NUID.0000.0133.2274 >> male >> d28 >> M6 >> >> NUID.0000.0133.2241 >> female >> d1 >> F7 >> >> NUID.0000.0133.2242 >> female >> d2 >> F7 >> >> NUID.0000.0133.2243 >> female >> d8 >> F7 >> >> NUID.0000.0133.2244 >> female >> d15 >> F7 >> >> NUID.0000.0133.2245 >> female >> d28 >> F7 >> >> NUID.0000.0133.2246 >> female >> d1 >> F8 >> >> NUID.0000.0133.2247 >> female >> d2 >> F8 >> >> NUID.0000.0133.2248 >> female >> d8 >> F8 >> >> NUID.0000.0133.2249 >> female >> d15 >> F8 >> >> NUID.0000.0133.2250 >> female >> d28 >> F8 >> >> >> Following is the code I am using: >> library("limma") >> rna <- read.table('qced_non_redundant_rma_data.txt', header=T, row.names=1, >> sep= "\t") >> annotation <- read.table('Test_Samples_Annotation.txt', header=T, >> row.names=1, sep= "\t") >> ## rearrange the annotations to conform to the RMA data >> annotation <- annotation[colnames(rna),] >> ## check that the annotations and data match >> if (sum(sort(rownames(annotation)) == sort(colnames(rna))) != >> length(rownames(annotation))) { >> stop ("ERROR: annotation rownames and RMA colnames do not match") >> } >> donor <- factor(annotation$donor) >> day <- factor(annotation$day) >> gender <- factor(annotation$gender) >> design <- model.matrix(~ donor + gender + day, data=annotation) >> fit <- lmFit(rna,design) >> fit2 <- eBayes(fit) >> write.fit(fit,file="effects.txt",digits=30, adjust="BH", >> method="separate",sep="\t") >> >> >> >> Following is my sessionInfo: >> > sessionInfo() >> R version 2.10.1 (2009-12-14) >> x86_64-unknown-linux-gnu >> locale: >> [1] LC_CTYPE=en_US.iso885915 LC_NUMERIC=C >> [3] LC_TIME=en_US.iso885915 LC_COLLATE=en_US.iso885915 >> [5] LC_MONETARY=C LC_MESSAGES=en_US.iso885915 >> [7] LC_PAPER=en_US.iso885915 LC_NAME=C >> [9] LC_ADDRESS=C LC_TELEPHONE=C >> [11] LC_MEASUREMENT=en_US.iso885915 LC_IDENTIFICATION=C >> attached base packages: >> [1] stats graphics grDevices utils datasets methods base >> other attached packages: >> [1] limma_3.2.3 >> loaded via a namespace (and not attached): >> [1] tools_2.10.1 >> [[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 > > > > ______________________________________________________________________ The information in this email is confidential and intend...{{dropped:4}}
ADD REPLY
0
Entering edit mode
Hi Gordon, I tried model.matrix(~gender+gender:donor+day). I am still running into the same error. In terms of what I want to get out of this analysis is as follows: 1. Does gene expression vary significantly across donors on the same day? Is the variation higher in males than in females? What are the least varying genes across donors? 2. Does gene expression vary significantly across days for a given donor? Is the variation higher in males than in females? What are the least varying genes across days? 3. What are the least varying genes across donors, days and gender all taken together? (more like house keeping genes) Any help would be greatly appreciated. What I tried: R version 2.10.1 (2009-12-14) Copyright (C) 2009 The R Foundation for Statistical Computing ISBN 3-900051-07-0 R is free software and comes with ABSOLUTELY NO WARRANTY. You are welcome to redistribute it under certain conditions. Type 'license()' or 'licence()' for distribution details. Natural language support but running in an English locale R is a collaborative project with many contributors. Type 'contributors()' for more information and 'citation()' on how to cite R or R packages in publications. Type 'demo()' for some demos, 'help()' for on-line help, or 'help.start()' for an HTML browser interface to help. Type 'q()' to quit R. > library("limma") > rna <- read.table('qced_non_redundant_rma_data.txt', header=T, row.names=1, sep= "\t") > annotation <- read.table('Test_Samples_Annotation.txt', header=T, row.names=1, sep= "\t") > dim(rna) [1] 54675 36 > ## check that the annotations and data match > if (sum(sort(rownames(annotation)) == sort(colnames(rna))) != length(rownames(annotation))) { + stop ("ERROR: annotation rownames and RMA colnames do not match") + } > donor <- factor(annotation$donor) > day <- factor(annotation$day) > gender <- factor(annotation$gender) > design <- model.matrix(~ donor + gender:donor + day, data=annotation) > fit <- lmFit(rna,design) Coefficients not estimable: donorF2:gendermale donorF3:gendermale donorF5:gendermale donorF7:gendermale donorF8:gendermale donorM1:gendermale donorM4:gendermale donorM6:gendermale Warning message: Partial NA coefficients for 54675 probe(s) > fit2 <- eBayes(fit) Warning message: In ebayes(fit = fit, proportion = proportion, stdev.coef.lim = stdev.coef.lim) : Estimation of var.prior failed - set to default value > write.fit(fit,file="effects.txt",digits=30, adjust="BH", method="separate",sep="\t") Warning message: In ebayes(fit = fit, proportion = proportion, stdev.coef.lim = stdev.coef.lim) : Estimation of var.prior failed - set to default value > sessionInfo() R version 2.10.1 (2009-12-14) x86_64-unknown-linux-gnu locale: [1] LC_CTYPE=en_US.iso885915 LC_NUMERIC=C [3] LC_TIME=en_US.iso885915 LC_COLLATE=en_US.iso885915 [5] LC_MONETARY=C LC_MESSAGES=en_US.iso885915 [7] LC_PAPER=en_US.iso885915 LC_NAME=C [9] LC_ADDRESS=C LC_TELEPHONE=C [11] LC_MEASUREMENT=en_US.iso885915 LC_IDENTIFICATION=C attached base packages: [1] stats graphics grDevices utils datasets methods base other attached packages: [1] limma_3.2.3 > > Date: Wed, 8 Feb 2012 14:23:44 +1100 > From: smyth@wehi.EDU.AU > To: genome1976@hotmail.com > CC: naomi@stat.psu.edu; bioconductor@r-project.org > Subject: Re: [BioC] Coefficients not estimable > > Or perhaps as a nested fixed model > > model.matrix(~gender+gender:donor+day) > > Whether to fit donor as a fixed term or as a block really depends on what > questions you want to answer and, in particular, what you want to > interpret from a donor "main effect". > > Best wishes > Gordon > > --------------------------------------------- > Professor Gordon K Smyth, > Bioinformatics Division, > Walter and Eliza Hall Institute of Medical Research, > 1G Royal Parade, Parkville, Vic 3052, Australia. > smyth@wehi.edu.au > http://www.wehi.edu.au > http://www.statsci.org/smyth > > On Tue, 7 Feb 2012, Naomi Altman wrote: > > > The problem is that donor is nested within gender. The only way I am aware > > of to handle this in limma is to use donor as a block. > > > > Regards, > > Naomi > > > > > > At 11:24 AM 2/7/2012, somnath bandyopadhyay wrote: > > > > > >> Dear Gordon and limma users, > >> > >> I am trying to fit a linear model to assess variability in healthy donors > >> over days for both genders using Affymetrix microarray data. I am > >> interested in assessing variability arising from geneder, day of blood > >> donated and individual donors. When I tried to do a linear fit in using > >> limma, I am not able to assess the coefficients for gender.System is > >> computationally singular for the gender main effect. Is there any way to > >> assess the variability arising from all three main effects? Any help would > >> be greatly appreciated. > >> > >> Following is my target file: > >> > >> > >> > >> > >> > >> > >> > >> > >> Array_NUID > >> gender > >> day > >> donor > >> > >> NUID.0000.0133.2260 > >> male > >> d1 > >> M1 > >> > >> NUID.0000.0133.2261 > >> male > >> d2 > >> M1 > >> > >> NUID.0000.0133.2262 > >> male > >> d8 > >> M1 > >> > >> NUID.0000.0133.2230 > >> female > >> d1 > >> F2 > >> > >> NUID.0000.0133.6809 > >> female > >> d2 > >> F2 > >> > >> NUID.0000.0133.2251 > >> female > >> d8 > >> F2 > >> > >> NUID.0000.0133.2228 > >> female > >> d15 > >> F2 > >> > >> NUID.0000.0133.2229 > >> female > >> d28 > >> F2 > >> > >> NUID.0000.0133.2231 > >> female > >> d1 > >> F3 > >> > >> NUID.0000.0133.2232 > >> female > >> d2 > >> F3 > >> > >> NUID.0000.0133.2233 > >> female > >> d8 > >> F3 > >> > >> NUID.0000.0133.2234 > >> female > >> d15 > >> F3 > >> > >> NUID.0000.0133.2265 > >> male > >> d1 > >> M4 > >> > >> NUID.0000.0133.2267 > >> male > >> d8 > >> M4 > >> > >> NUID.0000.0133.2268 > >> male > >> d15 > >> M4 > >> > >> NUID.0000.0133.2269 > >> male > >> d28 > >> M4 > >> > >> NUID.0000.0133.2236 > >> female > >> d1 > >> F5 > >> > >> NUID.0000.0133.2237 > >> female > >> d2 > >> F5 > >> > >> NUID.0000.0133.2238 > >> female > >> d8 > >> F5 > >> > >> NUID.0000.0133.6810 > >> female > >> d15 > >> F5 > >> > >> NUID.0000.0133.2240 > >> female > >> d28 > >> F5 > >> > >> NUID.0000.0133.2270 > >> male > >> d1 > >> M6 > >> > >> NUID.0000.0133.2271 > >> male > >> d2 > >> M6 > >> > >> NUID.0000.0133.2272 > >> male > >> d8 > >> M6 > >> > >> NUID.0000.0133.2273 > >> male > >> d15 > >> M6 > >> > >> NUID.0000.0133.2274 > >> male > >> d28 > >> M6 > >> > >> NUID.0000.0133.2241 > >> female > >> d1 > >> F7 > >> > >> NUID.0000.0133.2242 > >> female > >> d2 > >> F7 > >> > >> NUID.0000.0133.2243 > >> female > >> d8 > >> F7 > >> > >> NUID.0000.0133.2244 > >> female > >> d15 > >> F7 > >> > >> NUID.0000.0133.2245 > >> female > >> d28 > >> F7 > >> > >> NUID.0000.0133.2246 > >> female > >> d1 > >> F8 > >> > >> NUID.0000.0133.2247 > >> female > >> d2 > >> F8 > >> > >> NUID.0000.0133.2248 > >> female > >> d8 > >> F8 > >> > >> NUID.0000.0133.2249 > >> female > >> d15 > >> F8 > >> > >> NUID.0000.0133.2250 > >> female > >> d28 > >> F8 > >> > >> > >> Following is the code I am using: > >> library("limma") > >> rna <- read.table('qced_non_redundant_rma_data.txt', header=T, row.names=1, > >> sep= "\t") > >> annotation <- read.table('Test_Samples_Annotation.txt', header=T, > >> row.names=1, sep= "\t") > >> ## rearrange the annotations to conform to the RMA data > >> annotation <- annotation[colnames(rna),] > >> ## check that the annotations and data match > >> if (sum(sort(rownames(annotation)) == sort(colnames(rna))) != > >> length(rownames(annotation))) { > >> stop ("ERROR: annotation rownames and RMA colnames do not match") > >> } > >> donor <- factor(annotation$donor) > >> day <- factor(annotation$day) > >> gender <- factor(annotation$gender) > >> design <- model.matrix(~ donor + gender + day, data=annotation) > >> fit <- lmFit(rna,design) > >> fit2 <- eBayes(fit) > >> write.fit(fit,file="effects.txt",digits=30, adjust="BH", > >> method="separate",sep="\t") > >> > >> > >> > >> Following is my sessionInfo: > >> > sessionInfo() > >> R version 2.10.1 (2009-12-14) > >> x86_64-unknown-linux-gnu > >> locale: > >> [1] LC_CTYPE=en_US.iso885915 LC_NUMERIC=C > >> [3] LC_TIME=en_US.iso885915 LC_COLLATE=en_US.iso885915 > >> [5] LC_MONETARY=C LC_MESSAGES=en_US.iso885915 > >> [7] LC_PAPER=en_US.iso885915 LC_NAME=C > >> [9] LC_ADDRESS=C LC_TELEPHONE=C > >> [11] LC_MEASUREMENT=en_US.iso885915 LC_IDENTIFICATION=C > >> attached base packages: > >> [1] stats graphics grDevices utils datasets methods base > >> other attached packages: > >> [1] limma_3.2.3 > >> loaded via a namespace (and not attached): > >> [1] tools_2.10.1 > >> [[alternative HTML version deleted]] > >> > >> _______________________________________________ > >> Bioconductor mailing list > >> Bioconductor@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 inte...{{dropped:9}}
ADD REPLY
0
Entering edit mode
Dear Somnath, You are asking questions about variances, rather than questions about means; and questions about differential variability rather than questions about differential expression. You cannot easily answer these questions using limma, or I think any other Bioconductor package. Have a look at the following paper for the sort of methods that are relevant for analysing variability: http://www.ncbi.nlm.nih.gov/pubmed/21655321 These are statistically advanced methods, and you will need to find a very experienced statistician that you can work with. Best wishes Gordon --------------------------------------------- Professor Gordon K Smyth, Bioinformatics Division, Walter and Eliza Hall Institute of Medical Research, 1G Royal Parade, Parkville, Vic 3052, Australia. Tel: (03) 9345 2326, Fax (03) 9347 0852, smyth at wehi.edu.au http://www.wehi.edu.au http://www.statsci.org/smyth On Wed, 8 Feb 2012, somnath bandyopadhyay wrote: > > Hi Gordon, > > I tried model.matrix(~gender+gender:donor+day). I am still running into the same error. > > In terms of what I want to get out of this analysis is as follows: > > 1. Does gene expression vary significantly across donors on the same > day? Is the variation higher in males than in females? What are the > least varying genes across donors? > 2. Does gene expression vary significantly across days for a given > donor? Is the variation higher in males than in females? What are the > least varying genes across days? > 3. What are the least varying genes across donors, days and gender all > taken together? (more like house keeping genes) > > Any help would be greatly appreciated. > > What I tried: > > R version 2.10.1 (2009-12-14) > Copyright (C) 2009 The R Foundation for Statistical Computing > ISBN 3-900051-07-0 > R is free software and comes with ABSOLUTELY NO WARRANTY. > You are welcome to redistribute it under certain conditions. > Type 'license()' or 'licence()' for distribution details. > Natural language support but running in an English locale > R is a collaborative project with many contributors. > Type 'contributors()' for more information and > 'citation()' on how to cite R or R packages in publications. > Type 'demo()' for some demos, 'help()' for on-line help, or > 'help.start()' for an HTML browser interface to help. > Type 'q()' to quit R. >> library("limma") >> rna <- read.table('qced_non_redundant_rma_data.txt', header=T, row.names=1, sep= "\t") >> annotation <- read.table('Test_Samples_Annotation.txt', header=T, row.names=1, sep= "\t") >> dim(rna) > [1] 54675 36 >> ## check that the annotations and data match >> if (sum(sort(rownames(annotation)) == sort(colnames(rna))) != length(rownames(annotation))) { > + stop ("ERROR: annotation rownames and RMA colnames do not match") > + } >> donor <- factor(annotation$donor) >> day <- factor(annotation$day) >> gender <- factor(annotation$gender) >> design <- model.matrix(~ donor + gender:donor + day, data=annotation) >> fit <- lmFit(rna,design) > Coefficients not estimable: donorF2:gendermale donorF3:gendermale donorF5:gendermale donorF7:gendermale donorF8:gendermale donorM1:gendermale donorM4:gendermale donorM6:gendermale > Warning message: > Partial NA coefficients for 54675 probe(s) >> fit2 <- eBayes(fit) > Warning message: > In ebayes(fit = fit, proportion = proportion, stdev.coef.lim = stdev.coef.lim) : > Estimation of var.prior failed - set to default value >> write.fit(fit,file="effects.txt",digits=30, adjust="BH", method="separate",sep="\t") > Warning message: > In ebayes(fit = fit, proportion = proportion, stdev.coef.lim = stdev.coef.lim) : > Estimation of var.prior failed - set to default value >> sessionInfo() > R version 2.10.1 (2009-12-14) > x86_64-unknown-linux-gnu > locale: > [1] LC_CTYPE=en_US.iso885915 LC_NUMERIC=C > [3] LC_TIME=en_US.iso885915 LC_COLLATE=en_US.iso885915 > [5] LC_MONETARY=C LC_MESSAGES=en_US.iso885915 > [7] LC_PAPER=en_US.iso885915 LC_NAME=C > [9] LC_ADDRESS=C LC_TELEPHONE=C > [11] LC_MEASUREMENT=en_US.iso885915 LC_IDENTIFICATION=C > attached base packages: > [1] stats graphics grDevices utils datasets methods base > other attached packages: > [1] limma_3.2.3 >> > > > >> Date: Wed, 8 Feb 2012 14:23:44 +1100 >> From: smyth at wehi.EDU.AU >> To: genome1976 at hotmail.com >> CC: naomi at stat.psu.edu; bioconductor at r-project.org >> Subject: Re: [BioC] Coefficients not estimable >> >> Or perhaps as a nested fixed model >> >> model.matrix(~gender+gender:donor+day) >> >> Whether to fit donor as a fixed term or as a block really depends on what >> questions you want to answer and, in particular, what you want to >> interpret from a donor "main effect". >> >> Best wishes >> Gordon >> >> --------------------------------------------- >> Professor Gordon K Smyth, >> Bioinformatics Division, >> Walter and Eliza Hall Institute of Medical Research, >> 1G Royal Parade, Parkville, Vic 3052, Australia. >> smyth at wehi.edu.au >> http://www.wehi.edu.au >> http://www.statsci.org/smyth >> >> On Tue, 7 Feb 2012, Naomi Altman wrote: >> >>> The problem is that donor is nested within gender. The only way I am aware >>> of to handle this in limma is to use donor as a block. >>> >>> Regards, >>> Naomi >>> >>> >>> At 11:24 AM 2/7/2012, somnath bandyopadhyay wrote: >>> >>> >>>> Dear Gordon and limma users, >>>> >>>> I am trying to fit a linear model to assess variability in healthy donors >>>> over days for both genders using Affymetrix microarray data. I am >>>> interested in assessing variability arising from geneder, day of blood >>>> donated and individual donors. When I tried to do a linear fit in using >>>> limma, I am not able to assess the coefficients for gender.System is >>>> computationally singular for the gender main effect. Is there any way to >>>> assess the variability arising from all three main effects? Any help would >>>> be greatly appreciated. >>>> >>>> Following is my target file: >>>> >>>> >>>> >>>> >>>> >>>> >>>> >>>> >>>> Array_NUID >>>> gender >>>> day >>>> donor >>>> >>>> NUID.0000.0133.2260 >>>> male >>>> d1 >>>> M1 >>>> >>>> NUID.0000.0133.2261 >>>> male >>>> d2 >>>> M1 >>>> >>>> NUID.0000.0133.2262 >>>> male >>>> d8 >>>> M1 >>>> >>>> NUID.0000.0133.2230 >>>> female >>>> d1 >>>> F2 >>>> >>>> NUID.0000.0133.6809 >>>> female >>>> d2 >>>> F2 >>>> >>>> NUID.0000.0133.2251 >>>> female >>>> d8 >>>> F2 >>>> >>>> NUID.0000.0133.2228 >>>> female >>>> d15 >>>> F2 >>>> >>>> NUID.0000.0133.2229 >>>> female >>>> d28 >>>> F2 >>>> >>>> NUID.0000.0133.2231 >>>> female >>>> d1 >>>> F3 >>>> >>>> NUID.0000.0133.2232 >>>> female >>>> d2 >>>> F3 >>>> >>>> NUID.0000.0133.2233 >>>> female >>>> d8 >>>> F3 >>>> >>>> NUID.0000.0133.2234 >>>> female >>>> d15 >>>> F3 >>>> >>>> NUID.0000.0133.2265 >>>> male >>>> d1 >>>> M4 >>>> >>>> NUID.0000.0133.2267 >>>> male >>>> d8 >>>> M4 >>>> >>>> NUID.0000.0133.2268 >>>> male >>>> d15 >>>> M4 >>>> >>>> NUID.0000.0133.2269 >>>> male >>>> d28 >>>> M4 >>>> >>>> NUID.0000.0133.2236 >>>> female >>>> d1 >>>> F5 >>>> >>>> NUID.0000.0133.2237 >>>> female >>>> d2 >>>> F5 >>>> >>>> NUID.0000.0133.2238 >>>> female >>>> d8 >>>> F5 >>>> >>>> NUID.0000.0133.6810 >>>> female >>>> d15 >>>> F5 >>>> >>>> NUID.0000.0133.2240 >>>> female >>>> d28 >>>> F5 >>>> >>>> NUID.0000.0133.2270 >>>> male >>>> d1 >>>> M6 >>>> >>>> NUID.0000.0133.2271 >>>> male >>>> d2 >>>> M6 >>>> >>>> NUID.0000.0133.2272 >>>> male >>>> d8 >>>> M6 >>>> >>>> NUID.0000.0133.2273 >>>> male >>>> d15 >>>> M6 >>>> >>>> NUID.0000.0133.2274 >>>> male >>>> d28 >>>> M6 >>>> >>>> NUID.0000.0133.2241 >>>> female >>>> d1 >>>> F7 >>>> >>>> NUID.0000.0133.2242 >>>> female >>>> d2 >>>> F7 >>>> >>>> NUID.0000.0133.2243 >>>> female >>>> d8 >>>> F7 >>>> >>>> NUID.0000.0133.2244 >>>> female >>>> d15 >>>> F7 >>>> >>>> NUID.0000.0133.2245 >>>> female >>>> d28 >>>> F7 >>>> >>>> NUID.0000.0133.2246 >>>> female >>>> d1 >>>> F8 >>>> >>>> NUID.0000.0133.2247 >>>> female >>>> d2 >>>> F8 >>>> >>>> NUID.0000.0133.2248 >>>> female >>>> d8 >>>> F8 >>>> >>>> NUID.0000.0133.2249 >>>> female >>>> d15 >>>> F8 >>>> >>>> NUID.0000.0133.2250 >>>> female >>>> d28 >>>> F8 >>>> >>>> >>>> Following is the code I am using: >>>> library("limma") >>>> rna <- read.table('qced_non_redundant_rma_data.txt', header=T, row.names=1, >>>> sep= "\t") >>>> annotation <- read.table('Test_Samples_Annotation.txt', header=T, >>>> row.names=1, sep= "\t") >>>> ## rearrange the annotations to conform to the RMA data >>>> annotation <- annotation[colnames(rna),] >>>> ## check that the annotations and data match >>>> if (sum(sort(rownames(annotation)) == sort(colnames(rna))) != >>>> length(rownames(annotation))) { >>>> stop ("ERROR: annotation rownames and RMA colnames do not match") >>>> } >>>> donor <- factor(annotation$donor) >>>> day <- factor(annotation$day) >>>> gender <- factor(annotation$gender) >>>> design <- model.matrix(~ donor + gender + day, data=annotation) >>>> fit <- lmFit(rna,design) >>>> fit2 <- eBayes(fit) >>>> write.fit(fit,file="effects.txt",digits=30, adjust="BH", >>>> method="separate",sep="\t") >>>> >>>> >>>> >>>> Following is my sessionInfo: >>>>> sessionInfo() >>>> R version 2.10.1 (2009-12-14) >>>> x86_64-unknown-linux-gnu >>>> locale: >>>> [1] LC_CTYPE=en_US.iso885915 LC_NUMERIC=C >>>> [3] LC_TIME=en_US.iso885915 LC_COLLATE=en_US.iso885915 >>>> [5] LC_MONETARY=C LC_MESSAGES=en_US.iso885915 >>>> [7] LC_PAPER=en_US.iso885915 LC_NAME=C >>>> [9] LC_ADDRESS=C LC_TELEPHONE=C >>>> [11] LC_MEASUREMENT=en_US.iso885915 LC_IDENTIFICATION=C >>>> attached base packages: >>>> [1] stats graphics grDevices utils datasets methods base >>>> other attached packages: >>>> [1] limma_3.2.3 >>>> loaded via a namespace (and not attached): >>>> [1] tools_2.10.1 >>>> [[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 >>> >>> >>> >>> >> >> ______________________________________________________________________ >> The information in this email is confidential and intended solely for the addressee. >> You must not disclose, forward, print or use it without the permission of the sender. >> ______________________________________________________________________ > ______________________________________________________________________ The information in this email is confidential and intend...{{dropped:4}}
ADD REPLY

Login before adding your answer.

Traffic: 853 users visited in the last hour
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