Atoms are made of a nucleus and electrons; nuclei are made of protons and neutrons, and these are made of quarks and gluons. We can ask if the quarks (and electrons) are themselves made of something even smaller, which we usually call preons [pree-on].
My model has 7 assumptions.
The model described above corresponds to what mathematicians call a Group; and since I only expect a handful of preons, this is a Finite Group. These are generally well understood, and the continuous symmetries I find can be explained in terms of their representation theory. Unfortunately the theory requires a good deal of background, and the representations handiest for displaying the symmetry obscure the possible physics content.
To study the symmetries I decided to treat sums of group elements as vectors and look at the ways matrix transformations could mix them up without changing the essential nature of the group. The matrix transformation turns a group element into a sum of various group elements. I required that the collection of sums of group elements you get after the transformation of the group elements obeys the same group multiplication rules as did the original set. Put simply, the transformation can mix things up but the group better still work. This turns out to be a pretty stiff requirement.