Inferential Framework for Regulatory Subunits of Biomolecular Complexes from Proteome Datasets Using a Grand Canonical Ensemble Approach

The intracellular environment is highly dynamic, with proteins regulating themselves according to different cell states to perform various biological functions. While the static structures of proteins can be inferred from their sequences, investigating dynamic protein complexes in a crowded cell remains challenging. Recent technological advances have revealed certain aspects of dynamic and transient interactions of protein complexes, yet the formation of these complexes regulated by the protein abundance differed in cell states remain largely unknown. In this study, we provided a framework to infer the subunits that regulate protein assemblages. By using coarse-grained molecular modeling and proteome datasets, we quantified the fluctuations in subunits’ abundance in a grand canonical ensemble. Surprisingly, only a few subunits can drive a shift in state of the assemblages by slightly varying their abundance relative to the particle fluctuations in the simulations. We propose a mechanism where only few subunits whose abundance are strictly regulated function as regulators to attract other subunits from their reservoir. For example, in the INO80 chromatin remodeling complex, we discovered that the assemblage formation largely depends on few subunits with a high interaction to abundance ratio. Moreover, complex’s composition and regulators also depend on the total volume fraction of the protein subunits. We developed an approach to predict key regulators of biomolecular complex formation linked to morphological changes in cellular states, which provides insights into chromatin remodeling driven by dynamic protein abundance across different cellular conditions.