AsthmaML: Identification of Gene Signature Profiles with Deep Learning

Overview

I trained deep neural networks in Microsoft Azure to recognize gene signature profiles of asthma, a disease that is historically underdiagonosed. Using a novel gene clustering algorithm for feature selection, my deep learning model achieved 95.8% accuracy on data from the Gene Expression Omnibus (GEO).

Poster

View a PDF version of the poster here.

Source Code

Soon to be open source on GitHub.