Interaction Map Frameworks Reveal Sequence-Encoded Function and Complexation in Coevolved Disordered Proteins

Intrinsically disordered proteins (IDPs) are difficult to classify because they lack stable structures and show little sequence conservation. We address this challenge using the co-evolved NCBD and CID pair, across ancestral and modern and species. CID is highly disordered and charged, while NCBD has greater secondary structure and lower charge. Using polymer-physics-based interaction maps derived from a coarse-grained Hamiltonian and analytical theory, we examine how sequence patterning encodes function. A charge-based interaction map, cleanly separates CID into two major groups but fails to classify NCBD, which depends more on non-charge patterning. We therefore developed several quantitative non-charged interaction maps, and all consistently classify NCBD. These non-charge metrics qualitatively describe the complexation between NCBD and CID, a process relevant to gene expression. Together, these new statistical-physics approaches provide a unified, sequence-level framework for classifying IDPs and understanding how their sequence encodes both function and binding.