Hello, I was hoping someone could explain what, exactly, the Intercept part of the results shows?

I have a two factor model. Three treatment groups (treatment 1, treatment 2, and control) in two species (species 1 and species 2). I built the model and ran the Wald test as follows:

dds <- DESeqDataSetFromMatrix(countData = countdata, colData = coldata, design = ~ Treatment + Species + Treatment:Species); ddstotal <- DESeq(dds); matrix(resultsNames(ddstreatmenteffect)) [,1]
[1,] "Intercept"
[2,] "Speciesspecies1vsspecies2"
[3,] "Treatmenttreatment1vscontrol"
[4,] "Treatmenttreatment2vscontrol" [5,] "Speciesspecies1.treatment1"
[6,] "Speciesspecies1.treatment2"

From what I understand the results are telling me that there is a gene expression difference between species 1 and species 2 without accounting for treatment state, a difference between treatment 1 and untreated control and treatment2 and untreated control without accounting for species.

Results 5 and 6, I interpret to mean that the difference in treatment response between treatment 1 and untreated control (and treatment 2 and untreated control) is driven by species 1, and species 2 shows less difference...is this correct?

My question also is what information does result 1 "Intercept" tell me?

resint <- results(ddstotal, name = 'Intercept') head(res_int) log2 fold change (MLE): Intercept Wald test p-value: Intercept DataFrame with 6 rows and 6 columns baseMean log2FoldChange lfcSE stat pvalue padj <numeric> <numeric> <numeric> <numeric> <numeric> <numeric> A1BG 4.2760483829696 3.32823269926868 1.37359187717572 2.42301425523275 0.0153923226613891 0.0161058740430429 A4GALT 1111.95587001312 10.1174235883407 0.216407793931249 46.7516599312263 0 0 A4GNT 53.8968615744269 3.81977837078875 0.807954813571837 4.72771287035488 2.2706296498025e-06 2.47638780652672e-06

The intercept means different things in different models, in this model it would be log2 of the mean normalized counts for the group that is the reference level for all factor variables.

From what I understand the results are telling me that there is a gene expression difference between species 1 and species 2 without accounting for treatment state, a difference between treatment 1 and untreated control and treatment2 and untreated control without accounting for species.

What you posted there are not results. What is posted there are the names you use to conduct every kind of comparison you can with your design of ~ Treatment + Species + Treatment:Species. You don't have results until you run 'results'.