Year: 2026
Type: preprint
Abstract: We model the spread of a sexually transmitted infection on a partnership net-work in which every individual has exactly two partners of the opposite sex andno pair shares more than one partnership. The partnership graph is a random 2-regular simple bipartite graph on 2N vertices, which breaks into separate loopsof people. Under certain transmission the infection travels all the way aroundwhichever loop contains the first case, and cannot leave it: each loop is a groupin which all would be infected together if the disease reached any one of them.We show that the central risk quantity—the probability that a named individualis infected by a single introduced case—depends decisively on how the networkis randomised. Under the Derangement Model, appropriate when partnershipsform in two organised rounds, this probability converges to 1/2; under the Uni-form Model, appropriate for organic partnership formation, it converges to 2/3.We give exact O(N )-running-time simulation algorithms for both models anda Monte Carlo study that verifies empirically the validity both algorithms. Amixed-degree variant, in which one man and one woman have one partner eachwhile everyone else has two, illustrates the robustness of the 2/3 limit underthe Uniform law. It also yields an exact finite-population result: those two one-partner (degree-one) individuals are always linked through a single chain ofpartnerships (are endpoints of the same path component), so if one is infectedthe other is eventually infected too.
Source: Research Square
Authors: Zhivko Nedev, Zdravko Botev
Language: English
Subfield: Applied Mathematics
Field: Mathematics
Domain: Physical Sciences
Open Access status: green
OpenAlex ID: W7226013823
