Across Southeast Asia, expanding and intensifying swine production sectors face major disease threats, yet data on livestock trade networks remains limited. This thesis aimed to characterise the live pig trade network in Cambodia to identify key points of vulnerability to pathogen introduction and dissemination, with a focus on swine Influenza A viruses (IAV). In Chapter 2, I systematically reviewed the literature on models applied to simulate contact networks among livestock. I identified seven model frameworks broadly classified as being mechanistic, statistical, or machine learning-based. Large variation in model applications, calibration to data, and validation approaches were observed. This chapter guided methodological choices made in subsequent chapters. In Chapters 3 and 4, I analysed data from a questionnaire-based, cross-sectional, network survey I co-conducted within four provinces in south-central Cambodia. In Chapter 3, the personal 'egocentric' networks of value chain actors (n=377) and their immediate swine trading partners (n=1,101) are described. Network analysis identified smallholder boar service providers, middlemen, and breeding farms as 'brokers' at a high risk of disease introduction and dissemination – having many inward and outward connections with producers. Breeding farms supplied pigs to all producer types, increasing opportunities for disease dissemination along the value chain. In Chapter 4, I employed a subclass of exponential random graph models (ERGMs), estimable from egocentric data, to dissect the factors relevant for network formation. Complete, sociocentric, networks were simulated from fitted ERGMs, and IAV transmission was modelled on them. Simulations revealed that epidemic probabilities were highest when seeding in breeding farms, which, in addition to boar-lenders became infected soonest. Breeding farms also had the highest node-level prevalence at epidemic stationarity highlighting them as potential targets for IAV virological surveillance. Collectively, this thesis sheds light on vulnerabilities in the Cambodian swine sector, presents opportunities for targeted disease control and surveillance, and demonstrates the utility of egocentric sampling methods paired with ERGMs for network characterisation in data-constrained settings.