Abstract: Biobank initiatives include millions phased genomes, outpacing the tools available to analyze them. Conventional genotype formats treat variants independently, ignoring the ancestry structure underlying genetic variation. Prior work introduced the Genotype Representation Graph, a graph that captures shared ancestry directly but relies on traversal algorithms that map poorly onto accelerators. We show that it can be expressed exactly as sparse linear algebra: under a topological ordering, the graph’s structure becomes triangular, so the core genotype computation reduces to a sparse triangular solve. Further, we decompose it into a sequence of matrix-vector multiplications, exposing parallelism enabling hardware-optimized primitives in place of hand-tuned kernels. The result is substantial, portable speedups across accelerator hardware, from cloud deployments to supercomputers.

Host: Prof. O. Schenk