University Honors

This project consists of using two types of single-cell datasets that are scanned from a developing mouse brain: Single-cell high-throughput chromatin conformation (scHi-C) and single-cell RNA sequencing (scRNA-seq). The goal of this project is to integrate the scHi-C data and scRNA-seq data by recreating the machine learning algorithm scGAD (single-cell Gene Associating Domain) from scratch and analyzing the accuracy of the recreated algorithm. We chose to recreate the scGAD algorithm from scratch in order to confirm that the research papers backing this algorithm are correct so that more research can be done into making new integration algorithms. I will explain step-by-step how we were able to accurately recreate the scGAD algorithm while also explaining the conceptual reasoning behind the decisions made at each step.