Integrated Circuits and Systems >

2025 , Vol. 2 >Issue 2: 67 - 80

DOI: https://doi.org/10.23919/ICS.2025.3552542

Co-Optimization for Large Language Models: Advances in Algorithm and Hardware

emGene: An Embodied LLM NGS Sequencer for Real-Time Precision Diagnostics

  • SHAOBO LUO 1, 3 ,
  • ALBERT YU 3 ,
  • ZHIYUAN XIE 3 ,
  • HONG HUANG 1 ,
  • MINGQIANG HUANG 1 ,
  • KAI LI 1 ,
  • YUK KAN PUN 3 ,
  • ZHIRU GUO 1 ,
  • SHUWEI LI 1 ,
  • YIMING ZHU 3 ,
  • CHANGHAI MAN 1 ,
  • HUIYUAN SUN 3 ,
  • TUNG-HAN CHANG 3 ,
  • ZIYI GUAN 1, 2 ,
  • QIYUAN ZHANG 1 ,
  • TINGTING WANG 1 ,
  • GUANQI PENG 1 ,
  • WENJUN CHEN 1 ,
  • YAN SUN 3 ,
  • GENGXIN CHEN 3 ,
  • MEI YAN 3 ,
  • HAO YU , 1
Expand
  • 1 School of Microelectronics, Southern University of Science and Technology, Shenzhen 518055, China
  • 2 Department of Electrical and Electronic Engineering, University of Hong Kong, Hong Kong
  • 3 GeneSense Technology Inc., Shanghai 201112, China
Hao Yu (e-mail: ).

Received date: 2025-01-01

  Revised date: 2025-03-02

  Accepted date: 2025-03-13

  Online published: 2025-10-22

Supported by

Shenzhen Science and Technology Program under Grant(KQTD20200820113051096)

Shenzhen Science and Technology Program under Grant(JCYJ20220818100217038)

Science and Technology Innovation Key R&D Program of Chongqing

Fold

Abstract

Precision medicine is revolutionizing global healthcare by enabling personalized diagnostics, disease prediction, and tailored treatment strategies. While the integration of genomics and data science holds immense potential to optimize precision therapeutic outcomes, a critical challenge lies in translating gene sequencing data into actionable insights for in vitro diagnostics. This bottleneck is largely attributed to the limitations of edge-side intelligent processing and automation. Despite advancements in gene sequencing technologies and bioinformatics tools, the workflow from sample collection to diagnostic report generation remains fragmented, inefficient, and lacks of intelligence. To address these challenges, we introduce an embodied LLM NGS sequencer on the edge for real-time, on-site smart genetic diagnostics. This instrument integrates a streamlined and comprehensive pipeline with deep learning networks for primary data analysis, machine learning for secondary data processing, and a large language model (LLM) optimized for tertiary data interpretation. The LLM is enhanced through quantization and compression, facilitating deployment on FPGA/GPU to accelerate diagnostic workflows. Experimental results showcased the superior performance by achieving a 13.72% increase in throughput, a 99.50% Q30%, and enable smart diagnostic on the edge with the performance up to 75 tokens/s. This work enables immediate, on-site DNA analysis, hence dramatically improving precision medicine’s accessibility and efficiency, and significantly advances diagnostic accuracy, automation, establishing a robust platform for AI-driven personalized medicine and setting a new benchmark for the future of healthcare delivery.

Key words： Gene sequencing; NGS sequencer; base-calling; deep learning neural network; large language model; genetic diagnosis; model quantization; model edge-side deployment

Cite this article

SHAOBO LUO , ALBERT YU , ZHIYUAN XIE , HONG HUANG , MINGQIANG HUANG , KAI LI , YUK KAN PUN , ZHIRU GUO , SHUWEI LI , YIMING ZHU , CHANGHAI MAN , HUIYUAN SUN , TUNG-HAN CHANG , ZIYI GUAN , QIYUAN ZHANG , TINGTING WANG , GUANQI PENG , WENJUN CHEN , YAN SUN , GENGXIN CHEN , MEI YAN , HAO YU . emGene: An Embodied LLM NGS Sequencer for Real-Time Precision Diagnostics[J]. Integrated Circuits and Systems, 2025 , 2(2) : 67 -80 . DOI: 10.23919/ICS.2025.3552542

References

Publishing order | Descend order by publishing year | Descend order by cited within
[1]
E. A. Ashley, “Towards precision medicine,” Nature Rev. Genet., vol. 17, no. 9, pp. 507-522, 2016.

[2]
D. Yadav, B. Patil-Takbhate, A. Khandagale, J. Bhawalkar, S. T. Ripathy, and P. Khopkar-Kale, “Next-generation sequencing transforming clinical practice and precision medicine,” Clin. Chem. Acta, vol. 551, pp. 1-9, 2023.

[3]
A. A. Friedman, A. Letai, D. E. Fisher, and K. T. Falherty, “Precision medicine for cancer with next-generation functional diagnostics,” Nature Rev. Genetics, vol. 15, pp. 747-756, 2015.

[4]
T.Wu et al., “A brief overview of ChatGPT: The history, status quo and potential future development,” IEEE/CAA J. Automatica Sinica, vol. 10, no. 5, pp. 1122-1136, May 2023.

[5]
X. Huang, H. Yu, X. Liu, Y. Jiang, M. Yan, and D. Wu, “A dual-mode large-arrayed CMOS ISFET sensor for accurate and high-throughput pH sensing in biomedical diagnosis,” IEEE Trans. Biomed. Eng., vol. 62, no. 9, pp. 2223-2233, Sep. 2015.

[6]
X. Huang, F. Wang, J. Guo, M. Yan, H. Yu, and K. S. Yeo, “A 64 × 64 1200fps CMOS ion-image sensor with suppressed fixed-pattern-noise for accurate high-throughput DNA sequencing,” in Proc. Symp.VLSI Circuits Dig. Tech. Papers, 2014, pp. 1-2.

[7]
Y. Jiang, X. Liu, X. Huang, Y. Shang, M. Yan, and H. Yu, “Lab-on- CMOS: A multi-modal CMOS sensor platform towards personalized DNA sequencing,” in Proc. IEEE Int. Symp. Circuits Syst., 2016, pp. 2226-2229.

[8]
X. Huang, Y. Jiang, Y. Shang, H. Yu, and L. Sun, “A CMOS THzsensing system towards label-free DNA sequencing,” in Proc. IEEE 11th Int. Conf. ASIC, 2015, pp. 1-4.

[9]
X. Liu, Q. Fan, Z. Chen, P. Wan, W. Mao, and H. Yu, “A review and analysis of current-mode biosensing front-end ICs for nanopore-based DNA sequencing,” Front. Electron., vol. 3, pp. 1-5, 2022.

[10]
X. Huang, J. Guo, X. Wang, M. Yan, Y. Kang, and H. Yu, “A contactimaging based microfluidic cytometer with machine-learning for singleframe super-resolution processing,” PLoS One, vol. 9, no. 8, pp. 1-9, 2014.

[11]
X. Liu et al., “A microfluidic cytometer for complete blood count with a 3.2-megapixel, 1.1- μm-pitch super-resolution image sensor in 65- nm BSI CMOS,” IEEE Trans. Biomed. Circuits Syst., vol. 11, no. 4, pp. 794-803, Aug. 2017.

[12]
Y. Jiang et al., “A nano-filter-integrated CMOS image sensor for fluorescent biomedical imaging,” in Proc. IEEE Biomed. Circuits Syst. Conf., 2018, pp. 1-4.

[13]
D. Jeon et al., “A 23-mW face recognition processor with mostly-read 5T memory in 40-nm CMOS,” IEEE J. Solid-State Circuits, vol. 52, no. 6, pp. 1628-1642, Jun. 2017.

[14]
C. A. Valencia, M. A. Pervaiz, A. Husami, Y. Qian, and K. Zhang, “Sanger sequencing principles, history, and landmarks,” in Next Generation Sequencing Technologies in Medical Genetics ( 1st ed. Berlin, Germany: Springer, 2013, pp. 3-11.

[15]
T.-Y. Ho, K. Chakrabarty, and P. Pop, “Digital microfluidic biochips: Recent research and emerging challenges,” in Proc. 9th IEEE/ACM/IFIP Int. Conf. Hardware/Softw. Codesign System Synth., 2011, pp. 335-343.

[16]
W. Li, “DRAGEN optimizes genomic variant detection,” Nature Genetics, vol. 56, no. 12, 2024, Art. no. 2589.

[17]
BGI, “ChatBGI,” 2024.[Online]. Available: https://chat.bgi.com.

[18]
N. R. Paland S. K. Pal, “A review on image segmentation techniques,” Pattern Recognit., vol. 26, no. 9, pp. 1277-1294, 1993.

[19]
A. Krizhevsky, I. Sutskever, and G. E. Hinton, “ImageNet classification with deep convolutional neural networks,” Commun. ACM, vol. 60, no. 6, pp. 84-90, 2017.

[20]
K. He, X. Zhang, S. Ren, and J. Sun, “Deep residual learning for image recognition,” in Proc. IEEE Conf. Comput. Vis. Pattern Recognit., 2016, pp. 770-778.

[21]
Y. Chen, Q. Zhang, T. Li, H. Yu, M. Tian, and C. Zhuo, “ANT-UNet: Accurate and noise-tolerant segmentation for pathology image processing,” ACM J. Emerg. Technol. Comput. Syst., vol. 18, no. 2, pp. 1-17, 2021.

[22]
Y. Lecun, L. Bottou, Y. Bengio, and P. Haffner, “Gradient-based learning applied to document recognition,” Proc. IEEE, vol. 86, no. 11, pp. 2278-2324, Nov. 1998.

[23]
V. Badrinarayanan, A. Kendall, and R. Cipolla, “SegNet: A deep convolutional encoder-decoder architecture for image segmentation,” IEEE Trans. Pattern Anal. Mach. Intell., vol. 39, no. 12, pp. 2481-2495, Dec. 2017.

[24]
M. Borkó, B. Bolgár, and P. Sarkozy, “Basecalling raw nanopore DNA sequencing reads using neural networks,” in Proc. 25th Minisymp., 2018, pp. 4-7.

[25]
M. A. Boemo, “DNAscent v2: Detecting replication forks in nanopore sequencing data with Deep learning,” BMC Genomic., vol. 22, no. 430, pp. 1-8, 2021.

[26]
W.-C. Kao, K. Stevens, and Y. S. Song, “BayesCall: A model-based base-calling algorithm for high-throughput short-read sequencing,” Genome Res., vol. 19, no. 10, pp. 1884-1895, 2009.

[27]
W.-C. Kao and Y. S. Song, “NaiveBayesCall: An efficient model-based base-calling algorithm for high-throughput sequencing,” J. Comput. Biol., vol. 18, no. 3, pp. 365-377, 2011.

[28]
T. Massingham and N. Goldman, “All your base: A fast and accurate probabilistic approach to base calling,” Genome Biol., vol. 13, no. 2, pp. 1-15, 2012.

[29]
C. J. Saunders et al., “Rapid whole-genome sequencing for genetic disease diagnosis in neonatal intensive care units,” Sci. Transl. Med., vol. 4, no. 154, pp. 1-27, 2012.

[30]
J. R. Priest et al., “Molecular diagnosis of long QT syndrome at 10 days of life by rapid whole genome sequencing,” Heart Rhythm, vol. 11, no. 10, pp. 1707-1713, 2014.

[31]
J. E. Petrikin, L. K. Willig, L. D. Smith, and S. F. Kingsmore, “Rapid whole genome sequencing and precision neonatology,” Seminars Perinatology, vol. 39, no. 8, pp. 623-631, 2015.

[32]
M. M. Clark et al., “Diagnosis of genetic diseases in seriously ill children by rapid whole-genome sequencing and automated phenotyping and interpretation,” Sci. Transl. Med., vol. 11, no. 489, 2019, Art. no. eaat6177.

[33]
S. D. Goenka et al., “Accelerated identification of disease-causing variants with ultra-rapid nanopore genome sequencing,” Nature Biotechnol., vol. 40, pp. 1035-1041, 2022.

[34]
P.Wang et al., “EasyEdit: An easy-to-use knowledge editing framework for large language models,” in Proc. 62nd Annu. Meeting Assoc. Comput. Linguistics (Volume 3:Syst. Demonstrations), Bangkok, Thailand, 2023, pp. 82-93.

[35]
K. Singhal et al., “Large language models encode clinical knowledge,” Nature, vol. 620, no. 7972, pp. 172-180, 2023.

[36]
Y. Li, Z. Li, R. Dan, S. Jiang, and Y. Zhang, “ChatDoctor: A medical chat model fine-tuned on a large language model meta-AI (LLaMA) using medical domain knowledge,” Cureus, vol. 15, no. 6, pp. 1-12, 2023.

[37]
H. Zhang et al., “HuatuoGPT, towards taming language model to be a doctor,” in Findings of the Association for Computational Linguistics: EMNLP, Singapore: Assoc. Comput. Linguistics, 2023, pp. 10859-10885.

[38]
S. Li et al., “BaseFormer: Transformer based base-caller for fast and accurate next generation sequencing,” in Proc. 44th Annu. Int. Conf. IEEE Eng. Med. Biol. Soc., 2022, pp. 463-466.

[39]
S. Luo et al., “Hierarchical DNN with heterogeneous computing enabled high-performance DNA sequencing,” in Proc. IEEE Asia Pacific Conf. Circuits Syst., 2022, pp. 35-40.

[40]
J. Long, E. Shelhamer, and T. Darrell, “Fully convolutional networks for semantic segmentation,” in Proc. IEEE Conf. Comput. Vis. Pattern Recognit., 2015, pp. 3431-3440.

[41]
O. Ronneberger, P. Fischer, and T. Brox, “U-Net: Convolutional networks for biomedical image segmentation,” in Proc. Med. Image Comput. Comput.-Assist. Intervention, 2015, pp. 234-241.

[42]
GeneSense Technology, “Achieving high-quality whole human exome sequencing with sentieon DNAScope and genesense StarSeq100 AI sequencer,” 2024.[Online]. Available: https://www.gene-sense.com/en/news/130.html

[43]
H. Li, “Aligning sequence reads, clone sequences and assembly contigs with BWA-MEM,” 2013, arXiv:1303.3997.

[44]
L. Niu et al., “Optimizing accuracy and efficiency in analyzing non- UMI Liquid biopsy datasets using the Sentieon ctDNA pipeline,” bioRxiv, pp. 1-14, 2024.

[45]
B. M. Good, E. L. Clarke, S. Loguercio, and A. I. Su, “Linking genes to diseases with a SNPedia-gene wiki mashup,” J. Biomed. Semantics, vol. 3, no. 1, pp. 1-10, 2012.

[46]
A. Pal, L. Kumar Umapathi, and M. Sankarasubbu, “MedMCQA: A large-scale multi-subject multi-choice dataset for medical domain question answering,” in Proc. Conf. Health, Inference, Learn., 2022, pp. 248-260.

[47]
S. Luo et al., “ChatDNA: An end-to-end genetics analysis pipeline with embedding databases and prompt learning,” in Proc. 7th IEEE Int. Conf. Artif. Intell. Circuits Syst., 2025, pp. 1-5.

[48]
P. D. Stenson et al., “The Human gene mutation database (HGMD): Optimizing its use in a clinical diagnostic or research setting,” Hum. Genetics, vol. 139, no. 10, pp. 1197-1207, 2020.

[49]
W. A. Qader, M. M. Ameen, and B. I. Ahmed, “An overview of bag of words; importance, implementation, applications, and challenges,” in Proc. Int. Eng. Conf., 2019, pp. 200-204.

[50]
K. W. Church, “Word2vec,” Nat. Lang. Process., vol. 23, no. 1, pp. 155-162, 2017.

[51]
J. Lee et al., “BioBERT: A pre-trained biomedical language representation model for biomedical text mining,” Bioinf., vol. 36, no. 4, pp. 1-7, 2019.

[52]
M. Ahmed, R. Seraj, and S. M. S. Islam, “The K-means algorithm: A comprehensive survey and performance evaluation,” Electron., vol. 9, pp. 1-12, 2020.

[53]
K. Li et al., “A 29.12 TOPS/W and 1.13 TOPS/mm2 NAS-optimized mixed-precision DNN accelerator with vector split- and-combination systolic in 28nm CMOS,” in Proc. IEEE Custom Integr. Circuits Conf., 2024, pp. 1-2.

[54]
Z. Guan, H. Huang, Y. Su, H. Huang, N. Wong, and H. Yu, “APTQ: Attention-aware post-training mixed-precision quantization for large language models,” in Proc. 61st IEEE/ACM Des. Automat. Conf., 2024, pp. 1-6.

[55]
K. Li et al., “A precision-scalable energy-efficient bit-split-andcombination vector systolic accelerator for NAS-optimized DNNs on edge,” in Proc. Des., Automat. Test Europe Conf. Exhib., 2022, pp. 730-735.

[56]
L. Dai et al., “An energy-efficient bit-split-and-combination systolic accelerator for NAS-based multi-precision convolution neural networks,” in Proc. 27th Asia South Pacific Des. Automat. Conf., 2022, pp. 448-453.

[57]
M. Huang et al., “EdgeLLM: A highly efficient CPU-FPGA heterogeneous edge accelerator for large language models,” IEEE Trans. Circuits Syst. I, Reg. Papers, early access, Mar. 06, 2025, doi: 10.1109/TCSI.2025.3546256.

[58]
M. Huang, J. Luo, C. Ding, Z. Wei, S. Huang, and H. Yu, “An integer-only and group-vector systolic accelerator for efficiently mapping vision transformer on edge,” IEEE Trans. Circuits Syst. I, Reg. Papers, vol. 70, no. 12, pp. 5289-5301, Dec. 2023.

[59]
D. I. Lou et al., “High-throughput DNA sequencing errors are reduced by orders of magnitude using circle sequencing,” Proc. Nat. Acad. Sci., vol. 110, no. 49, pp. 19872-19877, 2013.

[60]
B. Ewing and P. Green, “Base-calling of automated sequencer traces using Phred. II. Error probabilities,” Genome Res., vol. 8, no. 3, pp. 186-194, 1998.

Options
Outlines

/

〈 〉