Hello,

I am investigating drug resistance of several cancer cell lines to multiple drugs. My expression data is baseline (basal) gene expression of the cell lines. In my case, there is no control/test groups but resistant (R) cells and sensitive cells (S) to each drug.

I made a design df named "drug_resp" like this;

#Display design entitled as "drug_resp"

> drug_resp

          Afatinib  Drug_2  Drug_3  Drug_4  Drug_5 ... Drug_25
Cell_1        R       R       R       R       R          R
Cell_2        R       S       R       R       R          S
Cell_3        R       S       R       R       R          R
Cell_4        S       R       S       R       R          R
....
....
Cell_14       R       S       R       R       R          R
Cell_15       R       R       R       R       R          R
Cell_16       S       R       R       R       S          S
Cell_17       S       R       S       R       R          R

# Make data set for 1st drug and run DESeq2

> dataset_Afatinib <- DESeqDataSetFromMatrix(countData = expression_df,colData = drug_resp, ~Afatinib)

> dds <- DESeq(dataset_Afatinib,minReplicatesForReplace = Inf)

> Afatinib_R_S = results(dds, contrast=c("Afatinib", "R", "S"),cooksCutoff = T, independentFiltering=T, pAdjustMethod = "bonferroni")

With those steps, I believe I can identify differentially expressed genes in Afatinib Resistant cells versus Afatinib Sensitive Cells.

My aim is to do these steps for each drug. There is no issue on gene expression table, I have used tximport for salmon qualifications. My only concern is the design stage of the model.

I am not sure the way I am doing is correct. If the approach of mine is correct, I have to create a new "datasetDrug#" for each drug (1 to 25) then generate "dds" object for each dataset. That is not practically memory-wise but still manageable.

Can you help me to perform this analysis elegantly for each drug?