Is there a possibility of taking 3 ifferent DAB brown images and converting them to fluoresence composite images to be analysed in a multiplex manner interactively ?
While I would definitely recommend creating your own color vectors using single stain controls, you can roughly do what you are asking through the Image->Color->Color Deconvolution.
You would then take the DAB channel from the split images and merge them back together into a stack or composite image. Probably with one of the hematoxylin channels.
There are also threads on creating such images from whole slide images through QuPath elsewhere on the forum.
I strongly recommend reading up on color deconvolution so that your results are meaningful.