• DEEP: A Deep Learning Platform for Personal Cancer Genome Analysis

    About the Project

    Our recent paper in Cancer Research developed a deep learning algorithm to scan personal cancer genomes, enabling us to identify pathogenic somatic mutations and to accurately predict clinical outcomes. To meet the clinical demands, we are now teaming up to implement this algorithm for large-scale genome analysis by developing hardware infrastructure and software system. The project name DEEP is an abbreviation of Deep Estimation from Epigenome Prediction in our original publication.
    In this project, we will develop user friendly interfaces for data input and output, GPU-powered deep learning system to scan personal genomes, and HIPPA compliant storage system to host clinical cancer genomes. We will also integrate ACMG diagnostic guidelines in our system to standardize clinical interpretation of pathogenic mutations. Although initially developed for prostate cancer, the system can be broadly applied to any cancer genomes.

    Hiring: Software Engineer

    We are looking for talented software engineer to join us:
    • Proficiency in designing data structures, data modeling and schema design
    • Proficient in using Python, Pytorch and TensorFlow
    • 1+ Years experience developing software in a professional environment