Environmental factors play an important role in precision medicine and can benefit from emerging data sources/analytics. The information-theoretic analysis detects ~80,000 associations between 219 exposures and ~1,700 diseases by integrating geospatial exposure data and electronic health records. The discovered associations showed concordance with ~11,0000 co-indexed chemical-disease co-occurrences in PubMed (peak odds ratio=2.38). Network analysis discovered novel groups of exposures to diseases, which suggest both vulnerability patterns and synergistic effects of multiple exposures.

Learning Objective 1: To understand the sensitivity of mutual information analyses for association mining in large biomedical datasets

Learning Objective 2: To understand the sensitivity of network analyses for unveiling systems-level patterns in large biomedical networks


Jungwei Fan (Presenter)
University of Arizona

Samir Rachid Zaim, University of Arizona
Walter Piegorsch, University of Arizona
Jianrong Li, University of Arizona
Colleen Kenost, University of Arizona
Yves Lussier, University of Arizona

Presentation Materials: