Because of the way we had dissected the brain, we would expect some clusters to appear in multiple tissues. For example, our olfactory sample included the anterior-most part of the cortex and underlying tissue (intended to cover the anterior olfactory nucleus), and could overlap with the cortex samples. In order to allow clusters to merge across such boundaries, and in order to improve resolution in clustering, we pooled cells in broader categories, and split them by (mostly) neurotransmitter, as follows:RegionClassspinal cordGABAergic, glycinergicspinal cordglutamatergicpnsallhypothalamuspeptidergichindbrainGABAergic, glycinergichindbrainglutamatergicwhole brainneuroblastsforebrainGABAergicforebrainglutamatergicdi- and mesencephalonGABAergicdi- and mesencephalonglutamatergicwhole braingranule cellswhole braincholinergic and monoaminergicstriatummedium spiny neurons
