[BioC] Contrasts result interpretation

Yoong [guest] guest at bioconductor.org
Wed Mar 12 22:04:44 CET 2014

Hi Mike, 

I have a quick question on contrasts. I am little bit confused with the result that I am getting from the contrasts I am making. The reason is I have 2 timepoints (6h and 12h), 3 different genotypes (PI, Sal, and R), and two different temperatures (20C and 30C). I am interested in seeing what genes are upregulated and downregulated for all the genotypes from timepoint 6h to 12h at 30C versus 20C. I am not sure if I am achieving that from my R analysis. Can you please take a look at my R command lines and see if the results I am getting from my analysis are genes that are differentially expressed at 30C versus 20C from 6h to 12h over all the genotypes?

Many thanks,

 -- output of sessionInfo(): 

> DEAnatomy = read.table("C:/Users/yfy/Desktop/NewMicroscopyDE.txt", row.names =1 , header = T, sep = "\t")

> colData=data.frame(row.names = colnames(DEAnatomy),
+ temp=c("20C","20C","20C","20C","20C","20C","30C","30C","30C","30C","30C","30C","20C","20C","20C","20C","20C","20C","30C","30C","30C","30C","30C","30C","20C","20C","20C","20C","20C","20C","30C","30C","30C","30C","30C","30C"),
+ genotype=c("PI","PI","PI","PI","PI","PI","PI","PI","PI","PI","PI","PI","SAL","SAL","SAL","SAL","SAL","SAL","SAL","SAL","SAL","SAL","SAL","SAL","R","R","R","R","R","R","R","R","R","R","R","R"),
+ time=c("6","6","6","12","12","12","6","6","6","12","12","12","6","6","6","12","12","12","6","6","6","12","12","12","6","6","6","12","12","12","6","6","6","12","12","12"))

> library("DESeq2", lib.loc="C:/Users/yfy/Documents/R/win-library/3.0")

> dds = DESeqDataSetFromMatrix(countData = DEAnatomy, colData = colData, design = ~genotype+time+temp)

> colData(dds)$time = relevel(colData(dds)$ time, "6")

> dds2 = DESeq(dds, betaPrior = FALSE)

> resultsNames(dds2)
[1] "Intercept"           "genotype_RNAi_vs_PI"
[3] "genotype_SAL_vs_PI"  "time_12_vs_6"       
[5] "temp_30C_vs_20C"  

> resctrst2<-results(dds2, contrast=c("time","12","6"))

> res<-resctrst2[order(resctrst2$padj),]

> sig_results<-res[which(res$padj<.05),]

> mcols(resctrst2)
DataFrame with 6 rows and 2 columns
          type                     description
   <character>                     <character>
1 intermediate     the base mean over all rows
2      results  log2 fold change: time 12 vs 6
3      results    standard error: time 12 vs 6
4      results    Wald statistic: time 12 vs 6
5      results Wald test p-value: time 12 vs 6
6      results            BH adjusted p-values

Sent via the guest posting facility at bioconductor.org.

More information about the Bioconductor mailing list