reading microarray data in R

Question

[12021560-002]

hi I am reading microarray data through R. please pin point the mistake in the command.

fit = lmFit(gse21340eset, design_gse21340) Error in lmFit(gse21340eset, design_gse21340) : row dimension of design doesn't match column dimension of data object

[12021560-002]

yes no. of rows are not matching with no. of columns in objects and designs. could you please tell how to same these numbers?

[WouterDeCoster]

You should have the same samples as rows in your design and as columns in your countdata. How big is the difference? I assume the difference is 1? Did you import the file in R yourself? Does the file have a header? Have a look at colnames(gse21340eset).

You can inspect data with head(design_gse21340) to only see the upper part of a table.

[andrew.j.skelton73]

Can you amend your post to include the code that created those two objects? lmfit is a linear model function that fits a given design matrix to your input data. It's not possible to determine where you went wrong with your command without the whole script.

[12021560-002]

there is a huge difference. you can see.

dim(gse21340eset) [1] 7129 20 dim(design_gse21340) [1] 0 4

is there any error in command or file uploading section?

[TriS]

can you upload the code you used to create the design matrix? based on the code you posted there are no samples in it (dimensions 0 4) while it should be 20 rows and 4 cols

[12021560-002]

This is the script

library(limma)

groups = pData(phenoData(gse21340dat[[1]]))$description

groups=as.character(groups)

groups[groups=="disease state: control.family history: yes"]="control.FH"

groups[groups=="disease state: control.family history: no"]="control.noFH"

groups[groups=="disease state: replicate control.family history: yes"]="rep.control.FH"

groups[groups=="disease state: DM.family history: yes"]="T2D.FH"

f = factor(groups, levels=c("control.FH","control.noFH","rep.control.FH","T2D.FH"))

design_gse21340 = model.matrix(~0+f)

colnames(design_gse21340) = levels(f)

cont.matrix = makeContrasts(T2DvsCtrlNoFH=T2D.FH-control.noFH, levels=design_gse21340)

fit= lmFit(gse21340eset, design_gse21340)

fit2= contrasts.fit(fit, cont.matrix)

[andrew.j.skelton73]

How did you generate gse21340eset? Also what are the results of head(gse21340eset) and head(design_gse21340)?

[TriS]

also, what's the groups variable like? I feel that you need to go through each variable...and also I'd take a very good read from this before continuing any extra coding. https://bioconductor.org/packages/release/bioc/vignettes/limma/inst/doc/usersguide.pdf

[12021560-002]

i think there is something in design matrix.

head(gse21340eset) GSM533283.CEL.gz GSM533284.CEL.gz GSM533285.CEL.gz GSM533286.CEL.gz A28102_at 8.130513 8.231653 8.175021 8.345722 AB000114_at 5.727701 5.696820 6.137276 5.878570 AB000115_at 6.684548 6.609435 6.887094 6.798657 AB000220_at 6.778942 6.961036 7.292673 7.298336 AB000381_s_at 4.940158 4.902318 4.885254 5.380067 AB000409_at 7.085271 7.080442 7.316466 6.942676 GSM533287.CEL.gz GSM533288.CEL.gz GSM533289.CEL.gz GSM533290.CEL.gz A28102_at 8.959552 8.473907 8.091576 8.414724 AB000114_at 6.517263 6.397724 6.100754 6.380587 AB000115_at 7.227893 7.349000 6.512520 6.769398 AB000220_at 7.967540 7.889100 7.657726 7.140753 AB000381_s_at 4.929813 5.059748 4.954322 4.910361 AB000409_at 7.142482 7.159999 6.572813 6.925385 GSM533291.CEL.gz GSM533292.CEL.gz GSM533293.CEL.gz GSM533294.CEL.gz A28102_at 9.083397 8.226454 7.554394 7.633388 AB000114_at 7.133230 6.469792 5.671915 5.303192 AB000115_at 7.651276 6.185414 6.483042 6.126700 AB000220_at 8.189044 7.843182 6.871853 6.594244 AB000381_s_at 4.934274 4.863728 4.907313 5.061410 AB000409_at 8.136080 6.801106 7.083652 6.524353 GSM533295.CEL.gz GSM533296.CEL.gz GSM533297.CEL.gz GSM533298.CEL.gz A28102_at 7.593197 7.644181 7.649320 7.772923 AB000114_at 5.372787 5.555148 5.653253 5.850298 AB000115_at 6.146925 6.516065 6.856782 6.556847 AB000220_at 6.388915 7.206200 6.611537 6.265709 AB000381_s_at 4.983873 5.007165 4.979237 4.996343 AB000409_at 6.701313 7.172355 7.061211 7.046278 GSM533299.CEL.gz GSM533300.CEL.gz GSM533301.CEL.gz GSM533302.CEL.gz A28102_at 7.711471 7.600967 7.582791 7.878591 AB000114_at 5.645259 5.696322 5.669252 6.349654 AB000115_at 6.393122 6.335462 6.663389 7.203055 AB000220_at 7.042607 7.247957 7.316512 7.843391 AB000381_s_at 4.935860 5.153809 5.052956 4.975595 AB000409_at 7.195019 6.953443 7.246403 7.539877

head(design_gse21340)

control.FH control.noFH rep.control.FH T2D.FH

Answer 1

[WouterDeCoster]

Have a look at

dim(gse21340eset)

and

dim(design_gse21340)

The function (apparently, not familiar with it myself) expects the number of rows in your design to match the number of columns in your data object.