I am working with human blood samples in a DE genes analysis. Now we want to perform the deconvolution of the matrix of counts to see the different cell types. My idea is to follow the guidelines described here, here and here but with RNA-seq data. I found that the R package DeconRNASeq can do it, as explained here. But, as far as I understand, the matrix-signature for each type of cell is required.
My question is: Where can I get the signatures for usual cell types present in blood to be used with DeconRNASeq?