How would you structure a data pipeline that handles both high-throughput NGS data and sparse, noisy wet-lab assay results at BigHat Biosciences?
