Teaching Machine Learning for Oncogenicity Prediction Based on NGS Genomic Metadata

In this paper, a complete machine learning solution is designed, implemented optimized as well as tested to cover automatic pathology examination of the DNA sequencing results produced by Next Generation Sequencer (NGS). The primary goal is to predict oncogenicity for single cells with unknown DNA mutation based on genomic metadata. Obtaining the goal could lead the physicians to make diagnose at an earlier stage, the patients to get the results faster and undergo less burdensome treatment, moreover, the hospitals and clinics to carry out a medical examination in a less expensive way.