Targeting Cancer via Signaling Pathways: A Novel Approach to the Discovery of Gene CCDC191's Double-Agent Function using Differential Gene Expression, Heat Map Analyses through AI Deep Learning, and Mathematical Modeling

Annie Ostojic

I developed a novel and replicable process using an artificial intelligence deep learning model that can be used to analyze the functions of genes in various pathways; specifically, I created this process through the discovery of gene CCDC191's functionality and its relationship in an abnormal pathway found in certain breast cancer patients.

Over the past couple of years, my family has experienced multiple encounters with cancer, some unfortunately, ending in death. This inspired me to tackle cancer using an analytical, computational biology approach. A need exists to study gene functions in pathways to meet a changing medical industry of personalized medicine and cancer treatments relative to gene expression patterns. In my computational biology and bioinformatics research project, I utilized a public access database to study gene CCDC191. Scientists do not know much about this gene other than its location and coiled-coil 3D structure; its function is unknown. Through an artificial intelligence deep learning model and reverse engineering mathematical modeling, I created a general process for determining the function of genes with unknown functions. Specifically, I found that the gene CCDC191 is a double agent which means it can lead to either the loss of controlled cell death or uncontrolled cell growth. Both can contribute to cancer development depending upon the type of cancer or situation. This study presents new insights into gene CCDC191, and it provides a replicable methodology which incorporates AI deep learning image classification and reverse engineering mathematical modeling to determine gene functions in pathways and cancer connectedness.