High-Performance Computing in Bioinformatics

Abstract

Bioinformatics workloads increasingly demand high-performance computing capabilities that exceed standard workstation capacity, from whole-genome alignment at population scale to protein structure prediction across proteomes and molecular dynamics simulations of biomolecular systems, and HPC architectures optimised for biological computation accelerate these workloads, yet adoption varies across HPC approaches and bioinformatics contexts. We evaluated 214 HPC bioinformatics programmes across centres affiliated with Mediterranean Institute of Technology in Rome between 2018 and 2022, spanning five architectural categories: traditional CPU cluster bioinformatics pipelines, GPU-accelerated sequence analysis and ML inference, distributed cloud bioinformatics platforms, containerised workflow orchestration systems, and quantum computing exploration for biological problems. A Bioinformatics HPC Index (BHI) was constructed from five sub-scores -- throughput speedup over baseline, cost-efficiency per analysis, reproducibility of results across platforms, workload scalability from lab to population scale, and accessibility to non-HPC-expert users - with weights from regression against sustained adoption. BHI correlated with adoption at r = +0.84 and discriminated adopted from non-adopted architectures with an AUC of 0.882. GPU-accelerated scored highest (mean BHI 0.824), while quantum computing trailed at 0.598. Only 35.5 percent exceeded the 0.75 threshold. Throughput speedup carried the largest weight (beta = +0.278), followed by accessibility (beta = +0.230)