Modeling Homophily in Dynamic Networks with Application to HIV Molecular Surveillance
This paper describes a novel approach to modeling homphily, i.e. the tendency of nodes that share (or differ in) certain attributes to be linked; we consider dynamic networks in which nodes can be added over time but not removed. Our application is to HIV genetic linkage analysis that has been used to investigate HIV transmission dynamics. In this setting, two HIV sequences from different persons with HIV (PWH) are said to be linked if the genetic distance between these sequences is less than a given threshold. Such linkage suggests that that the nodes representing the two infected PWH, are close to each other in a transmission network; such proximity would imply that either one of the infected people directly transmitted the virus to the other or indirectly transmitted it through a small number of intermediaries. These viral genetic linkage networks are dynamic in the sense that, over time, a group or cluster of genetically linked viral sequences may increase in size as new people are infected by those in the cluster either directly or through intermediaries. Our approach makes use of a logistic model to describe homophily with regard to demographic and behavioral characteristics that is we investigate whether similarities (or differences) between PWH in these characteristics impacts the probability that their sequences are be linked. Such analyses provide information about HIV transmission dynamics within a population.
Code (0)
등록된 구현이 없습니다.
Similar Papers 제목 키워드 기반
Dynamic Homophily with Imperfect Recall: Modeling Resilience in Adversarial Networks
The purpose of this study is to investigate how homophily, memory constraints, and adversarial disruptions collectively shape the resilience and adaptability of complex networks. To achieve this, we develop a new framewo…
Network Formation and Dynamics Among Multi-LLMs
Social networks fundamentally shape human opinions, behaviors, and the dissemination of information. As large language models (LLMs) like GPT, Claude, and Llama increasingly integrate into social and professional setting…
Decision MakingGraph Neural Networks with Heterophily
Graph Neural Networks (GNNs) have proven to be useful for many different practical applications. However, many existing GNN models have implicitly assumed homophily among the nodes connected in the graph, and therefore h…
Homophily and infections: static and dynamic effects
We analyze the effect of homophily in the diffusion of a harmful state between two groups of agents that differ in immunization rates. Homophily has a very different impact on the steady state infection level (that is in…
Graph Neural Networks for Molecules
Graph neural networks (GNNs), which are capable of learning representations from graphical data, are naturally suitable for modeling molecular systems. This review introduces GNNs and their various applications for small…
Molecular Property PredictionProperty PredictionSelf-Supervised Learning