Open Access. Powered by Scholars. Published by Universities.®
Articles 1 - 1 of 1
Full-Text Articles in Immune System Diseases
An Application Of The Modifiable Areal Unit Problem: Optimizing Cluster Method Parameters To Produce Predictive Data For Hiv Outbreaks, Connor J. Chato, Art Fy Poon
An Application Of The Modifiable Areal Unit Problem: Optimizing Cluster Method Parameters To Produce Predictive Data For Hiv Outbreaks, Connor J. Chato, Art Fy Poon
Western Research Forum
Background
A popular approach to study HIV outbreaks is to cluster cases based on genetic similarity. However, there is no widely-used statistical criterion which optimizes the parameters for sequence-based clustering methods. The relationship between a cluster-defining similarity threshold and it’s associated set of clusters can be analogized to the aggregation level in the Modifiable Areal Unit Problem (MAUP).
Hypothesis
Based on the selection of aggregation level for study partitions in MAUP, we present a statistical framework to optimize the similarity threshold for pairwise distance algorithm TN93 (http://github.com/veg/tn93). We hypothesize that defining this threshold includes case connections such that …