A Design of Polygenic Risk Model with Deep Learning for Colorectal Cancer in Multiethnic Indonesians
Steven Amadeus, Tjeng Wawan Cenggoro, Arif Budiarto, Bens Pardamean · Procedia Computer Science · 2021
Recently, health management is emerging and attract attention to how to provide better prognostication and health management systems. The challenges in the prognostication are how to develop a model that can self-learn the prognostication features and how to get a high accuracy prediction. Prognostication in health disease involves SNPs which is a genetic marker. In this paper, we propose a polygenic risk model using deep learning: Transformer with self-attention mechanism and DeepLIFT. The use of these deep learning model allows us to predict the risk of colorectal cancer and see the correlation between SNPs.