Patchy colloids are promising candidates for building blocks in directed self-assembly, but large scale synthesis of colloids with controlled surface patterns remains challenging. One potential fabrication method is to self-assemble the surface patterns themselves, allowing complex morphologies to organize spontaneously. For this approach to be competitive, prediction and control of the pattern formation process are necessary. However, structure formation in many-body systems is fundamentally hard to understand, and new theoretical methods are needed. Here we present a theory for self-assembling pattern formation in multi-component systems on the surfaces of colloidal particles, formulated as an analytic technique that predicts morphologies directly from the interactions in an effective model. As a demonstration we formulate an isotropic model of alkanethiols on gold, a suggested system for directed self-assembly, and predict its morphologies and transitions as a function of the interaction parameters.