Reference free, ultra fast valid statistical inference for sequencing data and biological sequences

Stanford scientists have created a statistical framework for interpreting next generation sequencing data which obviates the need for sequence alignment references in the most common and fundamental problems in genomics. The technology includes a unifying statistical formulation and proposes a solution to problems like finding genetic DNA nomads within and across species and identifying differential RNA expression or splicing that is near optimal in its computational efficiency with model-free valid statistical inference. The invention includes a reproducible computational workflow and the inventors have demonstrated the algorithm's ability to make biological discoveries missed by reference-based approaches.