Gene co-expression networks are used to model gene interactions across conditions with microarray or RNA-seq data. The inherent noise in such data coupled with more complicated analyses creates a “reproducibility crisis” in many of their conclusions. We propose a pipeline that leverages multiple asthma microarray datasets and uses a novel combination of techniques to increase the reliability of co-expression analysis. We highlight several key dysregulations across asthma and healthy patients that we believe have higher confidence than classical analyses.

Authors:

Brandon Guo (Presenter)
Monta Vista High School

Abhinav Kaushik, Stanford University
Kari Nadeau, Stanford University