EUVIMEDEUVIMED entdecken
Crossref

Toward an Embedded SoC for Mobile DNA Sequencing Applications

Zhongpan Wu, Abel Beyene, Karim Hammad, Ebrahim Ghafar-Zadeh, Sebastian Magierowski

IEEE Transactions on Computational Biology and Bioinformatics · 2025

DOI: 10.1109/tcbbio.2025.3602886 ↗

Kurz gefasst

Fachartikel in der Zeitschrift IEEE Transactions on Computational Biology and Bioinformatics, 2025. Verfasst von Zhongpan Wu und 4 weiteren Autor:innen. Hauptthemen laut MEDLINE: Sequence Analysis, DNA, Computational Biology, High-Throughput Nucleotide Sequencing. Die Arbeit nennt 35 Literaturangaben. Gefördert von: GlobalFoundries for providing silicon fabrication, 22FDX university program, Natural Sciences and Engineering Research Council of Canada.

AbstractSchlagwörterBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

The development of small DNA sequencers offers a transformative opportunity for portable genomics. However, current systems lack integrated bioinformatics computing and instead rely on bulky external systems. To address this limitation, we propose an embedded System-on-Chip (SoC) architecture aimed at mobile DNA sequencing applications, representing an initial step towards a fully integrated bioinformatics solution. The proposed heterogeneous SoC leverages the open-source RISC-V architecture and incorporates two specialized bioinformatics accelerators: one for deep learning-based basecalling and the other for sequence alignment. An FPGA-realized basecalling accelerator achieves a nearly 2000× speedup over a standalone RISC-V core and demonstrates 11.5× and 1.2× higher energy efficiency than x86 CPUs and high-end GPUs, respectively, while maintaining an accuracy rate of 83.7%. A separate accelerator optimized for sequence comparison, offers 538× and 357× performance-per-joule boost relative to x86 CPUs and an existing state-of-the-art accelerator, respectively, while far exceeding other platforms. A preliminary ASIC prototype operates at 250 MHz with a power consumption of 34 mW, demonstrating significant potential for the proposed approach. Overall, the proposed SoC architecture presents a compelling first step towards embedded bioinformatics hardware for mobile DNA sequencing.

Schlagwörter

MeSH-Schlagwörter (MEDLINE)

Sequence Analysis, DNAComputational BiologyHigh-Throughput Nucleotide SequencingGenomicsHumansDeep Learning

Förderung

  • GlobalFoundries for providing silicon fabrication
  • 22FDX university program
  • Natural Sciences and Engineering Research Council of Canada
  • CMC Microsystems
  • Qualcomm Canada Inc

Zitieren

APA

Wu, Z., Beyene, A., Hammad, K., Ghafar-Zadeh, E., & Magierowski, S. (2025). Toward an Embedded SoC for Mobile DNA Sequencing Applications. IEEE Transactions on Computational Biology and Bioinformatics. https://doi.org/10.1109/tcbbio.2025.3602886

Vancouver

Wu Z, Beyene A, Hammad K, Ghafar-Zadeh E, Magierowski S. Toward an Embedded SoC for Mobile DNA Sequencing Applications. IEEE Transactions on Computational Biology and Bioinformatics. 2025. doi:10.1109/tcbbio.2025.3602886

BibTeX
@article{Wu2025,
  author = {Wu, Zhongpan and Beyene, Abel and Hammad, Karim and Ghafar-Zadeh, Ebrahim and Magierowski, Sebastian},
  title = {Toward an Embedded SoC for Mobile DNA Sequencing Applications},
  journal = {IEEE Transactions on Computational Biology and Bioinformatics},
  year = {2025},
  doi = {10.1109/tcbbio.2025.3602886}
}
Kostenlos starten →