Gaussian Processes to Predict Gene Expression in Differentiating Induced Pluripotent Stem Cells

This study aims to answer the question of whether Gaussian Processes (GPs) are appropriate for prediction of gene expression in differentiating induced pluripotent stem cells (IPSCs) or whether these newer, more flexible methods overfit the data compared to the traditional baseline methods like linear regression.

Running Instructions

git clone https://github.com/nkrishn9/GP-IPSCs.git
cd GP-IPSCs/
chmod -x run_analysis.sh
source run_analysis.sh

This code is intended to run on the Midway cluster (UChicago), where the data resides.

Figures and Preliminary Results

Checkout the report.pdf in the root directory for more figures and a full explanation of the results of this project.