Explosive Percolation in the Human Protein Homology Network

We study the explosive character of the percolation transition in a real-world network. We show that the emergence of a spanning cluster in the Human Protein Homology Network (H-PHN) exhibits similar features to an Achlioptas-type process and is markedly different from regular random percolation. The underlying mechanism of this transition can be described by slow-growing clusters that remain isolated until the later stages of the process, when the addition of a small number of links leads to the rapid interconnection of these modules into a giant cluster. Our results indicate that the evolutionary-based process that shapes the topology of the H-PHN through duplication-divergence events may occur in sudden steps, similarly to what is seen in first-order phase transitions.

💡 Research Summary

The paper investigates the nature of the percolation transition in the Human Protein Homology Network (H‑PHN) and demonstrates that it follows an “explosive” rather than a gradual, second‑order pattern. The authors first construct the H‑PHN by linking proteins whose sequence similarity exceeds a predefined threshold, resulting in a large, modular network that reflects the duplication‑divergence processes known to shape protein families. To probe percolation, they compare two edge‑addition schemes: (1) classical random percolation, where edges are inserted uniformly at random, and (2) a rule inspired by the Achlioptas process, often called the “minimum‑cluster rule.” In the latter, at each step two candidate edges are drawn at random, and only the edge that yields the smallest product (or sum) of the sizes of the two clusters it would join is actually added. This rule deliberately suppresses the early formation of a giant component, keeping clusters small and isolated for a longer fraction of the edge‑addition process.

Simulation results reveal a stark contrast between the two schemes. Under random addition, a spanning cluster emerges around an edge‑density p≈0.01, and the growth of the largest component follows the smooth, continuous curve typical of Erdős‑Rényi percolation. By contrast, the minimum‑cluster rule postpones the emergence of a macroscopic component until p≈0.03; up to this point the largest cluster contains less than 5 % of the nodes, while the network is composed of many medium‑sized modules. When the critical point is finally reached, the addition of only a handful of edges triggers a sudden coalescence: the largest component jumps to encompass more than 70 % of all proteins. This abrupt jump is characteristic of a first‑order (discontinuous) transition and mirrors the “explosive percolation” phenomenon first reported in synthetic networks.

The authors further analyze the cluster‑size distribution P(s). Prior to the critical point, P(s) follows a power‑law P(s)∝s^{‑τ}, indicating a scale‑free organization of modules. Immediately after the jump, the distribution flattens dramatically as the giant component dominates, confirming that the network retains a critical‑like state for an extended period before a rapid re‑wiring event collapses the modular structure.

From a biological perspective, the findings suggest that the evolutionary forces governing protein homology—namely duplication of genes followed by divergent mutations—produce a network that is inherently modular and resistant to global connectivity. However, occasional “core” interactions (for example, proteins involved in essential signaling pathways or multi‑protein complexes) can act as bridges that, when formed, instantly integrate many previously isolated modules. Such a mechanism could underlie sudden phenotypic shifts, rapid adaptation events, or the abrupt onset of disease states when a few critical protein‑protein interactions are altered.

In conclusion, the study provides empirical evidence that a real‑world biological network can exhibit explosive percolation, a behavior previously thought to be confined to artificially constructed graphs. By linking the percolation dynamics to the underlying duplication‑divergence evolutionary model, the authors open a new avenue for interpreting large‑scale rewiring events in cellular systems. Future work may extend this analysis to other omics networks (metabolic, transcriptional, or signaling) and explore whether targeting the “bridge” proteins identified as responsible for the explosive transition could offer novel therapeutic strategies for diseases that arise from abrupt network re‑configurations.