A framework for protein and membrane interactions

We introduce the BioBeta Framework, a meta-model for both protein-level and membrane-level interactions of living cells. This formalism aims to provide a formal setting where to encode, compare and merge models at different abstraction levels; in particular, higher-level (e.g. membrane) activities can be given a formal biological justification in terms of low-level (i.e., protein) interactions. A BioBeta specification provides a protein signature together a set of protein reactions, in the spirit of the kappa-calculus. Moreover, the specification describes when a protein configuration triggers one of the only two membrane interaction allowed, that is “pinch” and “fuse”. In this paper we define the syntax and semantics of BioBeta, analyse its properties, give it an interpretation as biobigraphical reactive systems, and discuss its expressivity by comparing with kappa-calculus and modelling significant examples. Notably, BioBeta has been designed after a bigraphical metamodel for the same purposes. Hence, each instance of the calculus corresponds to a bigraphical reactive system, and vice versa (almost). Therefore, we can inherith the rich theory of bigraphs, such as the automatic construction of labelled transition systems and behavioural congruences.

💡 Research Summary

The paper introduces the BioBeta Framework, a formal meta‑model designed to capture both protein‑level and membrane‑level interactions within living cells. BioBeta builds on the well‑established kappa‑calculus by providing a protein signature (defining protein types, binding sites, and interaction rules) and a set of reaction rules that describe binding, unbinding, catalysis, and other elementary protein events. The novel contribution lies in the conditional membrane operations: only two membrane transformations are allowed—“pinch” (membrane invagination) and “fuse” (membrane fusion). These operations are triggered when specific protein configurations arise, thereby giving a rigorous biological justification for higher‑level membrane dynamics in terms of low‑level protein chemistry.

The semantics of BioBeta are given in terms of bigraphical reactive systems. A bigraph simultaneously represents spatial nesting (the “place” graph) and connectivity (the “link” graph), making it a natural substrate for encoding both the compartmental structure of membranes and the interaction network of proteins. Each BioBeta specification can be translated into a bigraph, and conversely (up to minor technical differences) each bigraphical reactive system corresponds to a BioBeta model. This near‑bijection allows the authors to inherit the rich theory of bigraphs: automatic construction of labelled transition systems (LTS), definition of behavioural congruences, and the use of existing bigraph tools for simulation and verification.

The authors analyse the expressive power of BioBeta by comparing it with pure kappa‑calculus. At the protein level, BioBeta is shown to be at least as expressive as kappa, reproducing all standard protein interaction patterns. At the membrane level, the addition of pinch and fuse operators enables concise representation of complex cellular processes that would otherwise require ad‑hoc extensions. To demonstrate practicality, two biologically significant case studies are modelled: (1) clathrin‑mediated endocytosis, where a specific protein complex triggers a pinch operation that creates a vesicle; and (2) SNARE‑mediated vesicle fusion, where the formation of a SNARE complex enables a fuse operation that merges two membranes. In both examples, the same BioBeta specification simultaneously captures the protein reaction network and the resulting membrane topology changes.

A further contribution is the discussion of model integration and comparison. Because BioBeta provides a common formal language, independently developed protein‑membrane models can be merged by translating them into a shared BioBeta signature. Formal equivalence checking, simulation composition, and behavioural analysis become feasible using the underlying bigraph machinery. The paper also outlines implementation prospects: existing bigraph tools can be leveraged to generate LTSs automatically, perform bisimulation checks, and support model checking of temporal properties.

In summary, BioBeta offers a unified, mathematically rigorous framework that bridges the gap between low‑level protein interaction modelling and high‑level membrane dynamics. By grounding membrane transformations in protein configurations and exploiting the bigraphical semantics, the framework delivers both expressive modelling capabilities and a pathway to automated analysis, positioning it as a valuable addition to the toolbox of systems and synthetic biologists.