Hello, I want to perform differential expression analysis with DESEQ2 of my miRNA seq data. I want to include age, sex, height, gender, and the batch in the design matrix as a covariate. In our data, we know that there is a batch effect that we have checked with PCA plot, there is a total of 16 batches in the data. I am adding these batches in binary form in the design matrix. here is my design: dds <- DESeqDataSetFromMatrix(se, colData = coldata, design = ~fev1+age+gender+ics+height+batch1+batch2+batch3+batch4+batch7+batch8+batch9+batch10+batch11+batch12+batch13+batch14+batch15+batch16+batch17+batch18) I am getting error : Error in checkFullRank(modelMatrix) : the model matrix is not full rank, so the model cannot be fit as specified. One or more variables or interaction terms in the design formula are linear combinations of the others and must be removed. Please read the vignette section 'Model matrix not full rank': vignette('DESeq2')

1. My data has a total of 365 samples and 645 miRNA
2. I am adding each covariate to the design matrix by converting them to the factor. Is this correct ??
3. I want to look DE miRNAs in FEV1 condition (Yes or No)
4. In the design matrix, all variables are in the binary except age and height they are continuous.

Please help me how do I solve this problem??

Thanks!!

Usually, one would have a batch variable with however many levels as you had batches, and then add simply batch to the design. R will then create the appropriate matrix to control for all levels of batch. This is the example in the vignette for example.

Aside from this, finding the problem with your design is up to you, and I can't help you in any particular way as the maintainer of DESeq2 to figure our the collinearity in your experiment.

You have a large complex dataset, so it may take some investigation, and you may need to consult with a local statistician.