Work

A scalable distributed workflow for accelerating long reads self-correction

Year: 2025

Type: article

Abstract: Third-Generation Sequencing (TGS) technologies have transformed genomic research by enabling the extraction of longer nucleotide sequences (referred to as long reads ) and providing deeper insights into genome structure. However, long reads are often associated with high sequencing error rates, making their correction a major challenge in many computational genomic pipelines. In this paper, we introduce HyperC , a distributed workflow designed to accelerate the execution of existing long-read self-correction tools through a hybrid parallelization strategy. By combining MPI and OpenMP, our proposal efficiently scatters and executes tasks across a distributed computing system. Optimized input data handling further reduces I/O bottlenecks and maximizes resource utilization. To assess the effectiveness of HyperC , we integrated it with CONSENT , a high-performance correction module, and conducted extensive experiments on real-world sequencing datasets. The results show significant reductions in execution time and improved scalability compared to the standalone execution of CONSENT , establishing HyperC as a robust and practical solution for high-performance genomic analysis and population-scale studies.

Source: Future Generation Computer Systems

DOI: 10.1016/j.future.2025.108244

Authors: Riccardo Ceccaroni, Lorenzo Di Rocco, Umberto Ferraro Petrillo, Pierpaolo Brutti

Institution: Sapienza University of Rome

Language: English

Topic: Genomics and Phylogenetic Studies

Subfield: Molecular Biology

Field: Biochemistry, Genetics and Molecular Biology

Domain: Life Sciences

Open Access status: hybrid

Estimated APC paid: $3,000

Funders: Istituto Nazionale di Alta Matematica "Francesco Severi", Sapienza Università di Roma, Gruppo Nazionale per l'Analisi Matematica, la Probabilità e le loro Applicazioni, Gruppo Nazionale per il Calcolo Scientifico