Network motif

A network motif is a small, recurring subgraph pattern — typically 3-5 nodes — that appears in a network with a frequency significantly higher than would be expected in a random graph of the same size and degree distribution. Motifs were first identified in biological networks by Uri Alon and colleagues, who found that feed-forward loops and bifan motifs appear disproportionately often in gene regulatory and neural circuits. Each motif implements a specific computational function: feed-forward loops act as persistence detectors, rejecting transient inputs while responding to sustained ones. The motif framework treats network topology as a library of reusable circuit elements rather than a single global structure. Understanding which motifs dominate a network reveals what computational problems the network has been optimized to solve, connecting network topology to functional adaptation in a way that global statistics cannot.