I try to analyze a GEO data-set (illumina human ht-12 v4). But I can not find the control profile, also can't find the "detection pval" column. So which file should I use? The GEO  number is GSE93219. Thanks a lot. 

Unfortunately the researchers who deposited GSE93219 chose to upload their data in a rather minimalistic form. They uploaded a matrix of (background corrected) intensity values but nothing else. They didn't upload IDAT files, detection p-values or control probe information.

Unfortunately this means that you cannot apply the neqc() normalization method recommended for Illumina BeadChips in the limma User's Guide. It isn't possible to reproduce the authors' own analysis pipeline either, which requires detection p-values, so you could legitimately write to the authors and ask them to provide more information.

Since all you have to use are intensities, I would suggest using the normexp background correction method followed by quantile normalization. You can do that like this.

First, read, in the data:

> Intensities <- read.delim("GSE93219_non-normalized.txt.gz",row.names="ID_REF",skip=4,comment.char="",quote="",check.names=FALSE)
> head(Intensities)
             7677145074_A 7677145074_D 7677145074_F 7677145074_I 7677145074_J
ILMN_1762337   -2.1116250  -2.37869400    -1.513151    4.1281220    2.1620370
ILMN_2055271    3.8533870   8.51458600    -1.531050    6.1021130   -4.8840290
ILMN_1736007   -0.6954489   0.04154839     4.007586   -0.9355530    0.3217204
ILMN_2383229    0.1194849   0.42363610     3.039766    1.5672620   -1.3412360
ILMN_1806310    2.1144340  -1.60486700     5.863817   -0.7480759    7.8510020
ILMN_1779670    3.9005440   2.62810400     1.189531    0.1058746   -1.6286350
             7677145074_L 7677145100_B 7677145100_D 7677145100_E 7677145100_G
ILMN_1762337   -1.8716880    2.8382580    0.5438465   9.08820200    -1.136811
ILMN_2055271   -1.5607060    3.1012740    7.1532370  13.20197000    15.902690
ILMN_1736007   -6.6722180   -0.7286062   -0.3817440   3.20575500    -0.412476
ILMN_2383229    0.2424704   -1.8650650   -4.8001250  -1.24175500     1.938118
ILMN_1806310    8.4159860   -0.3495244    6.0778200  -4.98494900    -5.160500
ILMN_1779670    5.5388950   -5.7164070    8.6839370   0.03834331    -3.216622
             7677145100_J 7677145100_K 7677145100_L
ILMN_1762337     1.414906   -7.0116310   -1.5758850
ILMN_2055271     3.147137    0.2801241   -0.4535244
ILMN_1736007    -4.763370   -2.8862500   -2.7357590
ILMN_2383229    -3.012951   -0.7996641   -2.5958350
ILMN_1806310    -2.477314    1.2336840    0.0526173
ILMN_1779670    -1.621891    0.3306231    2.5410890
> x <- new("EListRaw")
> x$E <- as.matrix(Intensities)

Then background correct and quantile normalize:

> xb <- backgroundCorrect(x,method="normexp",offset=16)
> y <- normalizeBetweenArrays(xb,method="quantile")

Then you can proceed with a DE analysis. The targets information is as follows:

> targets
             Donor Subset     Status
7677145074_A   106   Th2A Stimulated
7677145074_D   149   Th2A Stimulated
7677145074_F   106    Th1       Ctrl
7677145074_I   106    Th2       Ctrl
7677145074_J   106   Th2A       Ctrl
7677145074_L   149    Th1       Ctrl
7677145100_B    21    Th1 Stimulated
7677145100_D   149   Th17 Stimulated
7677145100_E   149    Th2 Stimulated
7677145100_G   149    Th1 Stimulated
7677145100_J    21   Th17 Stimulated
7677145100_K   149   Th2A       Ctrl
7677145100_L   149    Th2       Ctrl