Overview
Biology
MEK1 phosphorylates ERK2 on its Thr-X-Tyr motif in the
cytoplasm. Phospho-ERK then imports through nuclear pore
complexes; nuclear phosphatases reset it. MEK is active-site
limited — one ERK at a time.
Bigraph encoding
- Place graph:
Cell ⊃ Cytoplasm ⊃ {Nucleus, ERLumen, MEK, ERK, pERK}. - Link graph: one shared edge per MEK·pERK complex.
- Sorts:
Cell,Compartment,MEK,ERK,pERK+ kind
tags Cytoplasm, Nucleus, ERLumen.
Rules (Gillespie SSA, propensity = k × |matches|)
phosphorylate(k = 2.0): co-located free MEK + free ERK bind via a fresh edge.dissociate(k = 0.5): MEK·pERK → free MEK + free pERK.dephosphorylate(k = 0.4): nuclear pERK → ERK (DUSP-style reset).translocate_erk_in / out(k = 1.0 each): free ERK between cytoplasm and child compartment.translocate_perk_in(k = 2.0): free pERK, cytoplasm → nucleus (active import).translocate_perk_out(k = 0.1): free pERK, nucleus → cytoplasm (slow leak).
The 20-fold in/out asymmetry drives nuclear pERK accumulation
— the signal itself.
Why bigraphs?
Compartment-only models capture *where*; reaction-network models
capture *what's bound*. Bigraphs unify both — a single redex
constrains place AND link in one step (e.g. phosphorylate
requires co-location AND no prior bond). See
references/brs_mapk.md for citations.
Initial bigraph (place + link graph)
Substrate populations through time
Per-compartment substrate counts under Gillespie SSA, with rule-firing ticks above the curves.
Structural snapshots
Cell cartoons at evenly spaced points in the run — MEK in the cytoplasm, ERK / pERK distributed across compartments, MEK·pERK bonds drawn as paired green parallel lines.