pith. sign in

arxiv: 2605.25388 · v1 · pith:KXGH73A2new · submitted 2026-05-25 · 💻 cs.LG · q-bio.QM

ViroBench: Benchmarking Nucleotide Foundation Models on Viral Genomics Tasks

Dongxin Ye , Fang Hu , Han Hu , Shu Hu , Yang Tan , Wanli Ouyang , Stan Z. Li , Jie Cui
show 1 more author
Nanqing Dong
This is my paper

Pith reviewed 2026-06-29 22:28 UTC · model grok-4.3

classification 💻 cs.LG q-bio.QM
keywords ViroBenchnucleotide foundation modelsviral genomicsbiosecurity riskphylogenetic shifttemporal shiftpretraining diversitybenchmark
0
0 comments X

The pith

Nucleotide foundation models degrade on phylogenetic and temporal shifts in viruses and decouple likelihood from function in generations.

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The paper presents ViroBench as a benchmark for nucleotide foundation models in viral genomics. It evaluates 66 models on 18 scenarios in 4 task types covering biological understanding and biosecurity risks. The results indicate that models degrade in performance when facing new virus lineages or time periods. Generation tasks show that sequences with high statistical likelihood often lack biological function, indicating potential risks. A key finding is that training on diverse virus types improves performance more than increasing model size, with a small model gaining 67.5 percent over its base.

Core claim

ViroBench demonstrates that NFMs exhibit a performance degradation in biological understanding under phylogenetic and temporal shifts, indicating weak extrapolation capabilities. Generation tasks reveal a decoupling between statistical likelihood and biological functional validity, posing latent biosecurity risks. Controlled ablation studies reveal that taxonomic diversity in pretraining data outweighs parameter scale, with a lightweight baseline achieving a 67.5% performance gain over its original model.

What carries the argument

ViroBench benchmark with 18 scenarios across 4 task types assessing biological understanding and latent biosecurity risk in viral nucleotide models.

Load-bearing premise

The 18 scenarios and 4 task types in ViroBench sufficiently represent the critical dimensions of biological understanding and latent biosecurity risk for viral genomics applications.

What would settle it

Observing an NFM that maintains performance across phylogenetic and temporal shifts or where high-likelihood generations consistently match functional validity would falsify the degradation and decoupling claims.

Figures

Figures reproduced from arXiv: 2605.25388 by Dongxin Ye, Fang Hu, Han Hu, Jie Cui, Nanqing Dong, Shu Hu, Stan Z. Li, Wanli Ouyang, Yang Tan.

Figure 1
Figure 1. Figure 1: Overview of ViroBench. (a) Data Composition: 58,314 sequences featuring hierarchical taxonomy, host categories, [PITH_FULL_IMAGE:figures/full_fig_p003_1.png] view at source ↗
Figure 2
Figure 2. Figure 2: Experimental results. (a) Family-level confusion matrix (left) and phylogenetic tree (right) for AIDO.DNA-7B, vi [PITH_FULL_IMAGE:figures/full_fig_p007_2.png] view at source ↗
Figure 3
Figure 3. Figure 3: Distribution of the model’s maximum predicted [PITH_FULL_IMAGE:figures/full_fig_p012_3.png] view at source ↗
Figure 4
Figure 4. Figure 4: Detailed configuration of the baseline 1D-ResNet architecture. The schematic illustrates the end-to-end processing [PITH_FULL_IMAGE:figures/full_fig_p016_4.png] view at source ↗
Figure 5
Figure 5. Figure 5: Performance evaluation and phylogenetic analysis of viral family classification across four models. The left column [PITH_FULL_IMAGE:figures/full_fig_p018_5.png] view at source ↗
Figure 6
Figure 6. Figure 6: Multi-dimensional diagnostic analysis of performance disparities ( [PITH_FULL_IMAGE:figures/full_fig_p019_6.png] view at source ↗
Figure 7
Figure 7. Figure 7: The t-SNE dimensionality reduction distributions of viral genome embedding vectors generated by four models [PITH_FULL_IMAGE:figures/full_fig_p020_7.png] view at source ↗
Figure 8
Figure 8. Figure 8: The monthly aggregate time trend of Evo2-40B. The top chart shows the BPB (left axis, blue) and K-mer JSD (right [PITH_FULL_IMAGE:figures/full_fig_p023_8.png] view at source ↗
Figure 9
Figure 9. Figure 9: The generated and ground sequence structures of the top three targets in each model group are overlaid. Each panel [PITH_FULL_IMAGE:figures/full_fig_p024_9.png] view at source ↗
Figure 10
Figure 10. Figure 10: Structural confidence and fold similarity between generated and ground-truth proteins. (a) Distributions of per-target [PITH_FULL_IMAGE:figures/full_fig_p025_10.png] view at source ↗
Figure 11
Figure 11. Figure 11: Joint analysis of classification and generation performance. Each panel plots macro-F1 (x-axis) against a generation [PITH_FULL_IMAGE:figures/full_fig_p026_11.png] view at source ↗
Figure 12
Figure 12. Figure 12: Confusion matrices for hierarchical taxonomic and host classification on 89 Nipah virus genomes across five [PITH_FULL_IMAGE:figures/full_fig_p027_12.png] view at source ↗
Figure 13
Figure 13. Figure 13: Sliding-window PPL profiles along the Nipah virus genome (AF212302.2). The top track shows annotated CDS regions, [PITH_FULL_IMAGE:figures/full_fig_p027_13.png] view at source ↗
Figure 14
Figure 14. Figure 14: Distribution of Macro-F1 scores across all classifi [PITH_FULL_IMAGE:figures/full_fig_p028_14.png] view at source ↗
Figure 15
Figure 15. Figure 15: Mean BPB across genome-length buckets (short, [PITH_FULL_IMAGE:figures/full_fig_p029_15.png] view at source ↗
read the original abstract

Nucleotide sequences constitute the fundamental genetic basis of biological systems, rendering viral genomic analysis critical for biomedical advancement. Despite progress in biological foundation models, specifically nucleotide foundation models (NFMs), the field lacks a unified standard for viral genomics to facilitate community development and enforce biosecurity constraints. To address this, we introduce ViroBench, the first comprehensive and large-scale benchmark specifically designed for NFMs in viral settings. ViroBench evaluates models across two critical dimensions: biological understanding and latent biosecurity risk, covering 18 diverse scenarios within 4 task types. Extensive evaluation of 66 NFMs across diverse architectures yields three critical conclusions. Firstly, NFMs exhibit a performance degradation in biological understanding under phylogenetic and temporal shifts, indicating weak extrapolation capabilities. Secondly, generation tasks reveal a decoupling between statistical likelihood and biological functional validity, posing latent biosecurity risks. Thirdly, controlled ablation studies reveal that taxonomic diversity in pretraining data outweighs parameter scale. Specifically, a lightweight baseline trained on diverse data achieves a 67.5% performance gain over its original model. Overall, ViroBench provides interpretable, diagnostic evaluations and a reproducible measurement framework for future research on viral nucleotide foundation models. The datasets and code are publicly available at https://github.com/QIANJINYDX/ViroBench.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Referee Report

2 major / 0 minor

Summary. The paper introduces ViroBench, the first large-scale benchmark for nucleotide foundation models (NFMs) on viral genomics. It evaluates 66 NFMs across 18 scenarios in 4 task types spanning biological understanding and latent biosecurity risk. The central claims are that NFMs show performance degradation under phylogenetic and temporal shifts, that generation tasks exhibit decoupling between statistical likelihood and functional validity (raising biosecurity concerns), and that taxonomic diversity in pretraining data outweighs parameter scale, evidenced by a lightweight baseline achieving a 67.5% performance gain. Datasets and code are released publicly.

Significance. If the benchmark tasks and evaluation protocols are sound, ViroBench would fill a clear gap by providing a reproducible, diagnostic framework for NFMs in viral settings. The public release of datasets and code is a concrete strength that supports community follow-up. The three conclusions, if substantiated, would usefully highlight generalization limits and safety considerations for future NFM development.

major comments (2)
  1. [Abstract] Abstract: the three conclusions are presented as resulting from 'extensive evaluation' of 66 models, yet the text supplies no description of the 18 scenarios, the four task types, the metrics, statistical tests, or error analysis. This absence is load-bearing because the degradation, decoupling, and diversity-over-scale claims cannot be assessed without those details.
  2. [Abstract] Abstract: the 67.5% performance gain for the lightweight baseline is stated without identifying the original model, the exact metric, the ablation controls, or the composition of the 'diverse data,' preventing evaluation of whether taxonomic diversity truly outweighs scale.

Simulated Author's Rebuttal

2 responses · 0 unresolved

We thank the referee for their constructive feedback and for highlighting areas where the abstract could better support the manuscript's claims. We address each major comment below.

read point-by-point responses

  1. Referee: [Abstract] Abstract: the three conclusions are presented as resulting from 'extensive evaluation' of 66 models, yet the text supplies no description of the 18 scenarios, the four task types, the metrics, statistical tests, or error analysis. This absence is load-bearing because the degradation, decoupling, and diversity-over-scale claims cannot be assessed without those details.

    Authors: We agree the abstract is concise and omits these specifics. The main text details the 18 scenarios and four task types in Section 3, the metrics and evaluation protocols (including statistical tests) in Section 4, and error analysis in Section 5. To improve accessibility, we will revise the abstract to briefly name the task types and note the use of standard metrics with significance testing. revision: yes

  2. Referee: [Abstract] Abstract: the 67.5% performance gain for the lightweight baseline is stated without identifying the original model, the exact metric, the ablation controls, or the composition of the 'diverse data,' preventing evaluation of whether taxonomic diversity truly outweighs scale.

    Authors: The ablation details—including the original model, exact metric, controls, and diverse data composition—are reported in Section 5.3. We acknowledge the abstract lacks this context and will revise it to specify the model, metric, and key elements of the taxonomic diversity. revision: yes

Circularity Check

0 steps flagged

Empirical benchmark with no circular derivations

full rationale

This is a pure empirical benchmark paper that introduces ViroBench, evaluates 66 NFMs on 18 scenarios across 4 task types, and reports direct performance measurements. All three conclusions (degradation under shifts, likelihood-functional decoupling, and taxonomic diversity outweighing scale) are presented as observations from those measurements and ablations, with no equations, fitted parameters renamed as predictions, self-definitional steps, or load-bearing self-citations. The derivation chain is empty; results are self-contained as reported benchmark outcomes.

Axiom & Free-Parameter Ledger

0 free parameters · 0 axioms · 0 invented entities

No free parameters, axioms, or invented entities are introduced; the contribution is an empirical benchmark and evaluation framework relying on standard machine learning evaluation practices.

pith-pipeline@v0.9.1-grok · 5784 in / 1100 out tokens · 27150 ms · 2026-06-29T22:28:31.808700+00:00 · methodology

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Reference graph

Works this paper leans on

67 extracted references · 4 canonical work pages · 1 internal anchor

  1. [1]

    Josh Abramson, Jonas Adler, Jack Dunger, Richard Evans, Tim Green, Alexander Pritzel, Olaf Ronneberger, Lindsay Willmore, Andrew J Ballard, Joshua Bambrick, et al. 2024. Accurate structure prediction of biomolecular interactions with AlphaFold 3.Nature630, 8016 (2024), 493–500

    2024

  2. [2]

    Manato Akiyama and Yasubumi Sakakibara. 2022. Informative RNA base em- bedding for RNA structural alignment and clustering by deep representation learning.NAR genomics and bioinformatics4, 1 (2022), lqac012

    2022

  3. [3]

    P Balakrishnan, A Anny Leema, V Dhivya Shree, C Mohammad Saad, and A Mohan Babu. 2025. Gene-LLMs: a comprehensive survey of transformer-based genomic language models for regulatory and clinical genomics.Frontiers in Genetics16 (2025), 1634882

    2025

  4. [4]

    Dennis A Benson, Mark Cavanaugh, Karen Clark, Ilene Karsch-Mizrachi, David J Lipman, James Ostell, and Eric W Sayers. 2012. GenBank.Nucleic acids research 41, D1 (2012), D36–D42

    2012

  5. [5]

    Sam Boshar, Benjamin Evans, Ziqi Tang, Armand Picard, Yanis Adel, Franziska K Lorbeer, Chandana Rajesh, Tristan Karch, Shawn Sidbon, David Emms, et al

  6. [6]

    A foundational model for joint sequence-function multi-species modeling at scale for long-range genomic prediction.bioRxiv(2025), 2025–12

    2025

  7. [7]

    Sebastian Bowyer, David J Allen, and Nicholas Furnham. 2025. Unveiling the ghost: machine learning’s impact on the landscape of virology.Journal of General Virology106, 1 (2025), 002067

    2025

  8. [8]

    Garyk Brixi, Matthew G Durrant, Jerome Ku, Mohsen Naghipourfar, Michael Poli, Gwanggyu Sun, Greg Brockman, Daniel Chang, Alison Fanton, Gabriel A Gonzalez, et al. 2026. Genome modelling and design across all domains of life with Evo 2.Nature652, 8112 (2026), 1349–1361

    2026

  9. [9]

    Christiam Camacho, George Coulouris, Vahram Avagyan, Ning Ma, Jason Pa- padopoulos, Kevin Bealer, and Thomas L Madden. 2009. BLAST+: architecture and applications.BMC bioinformatics10, 1 (2009), 421

    2009

  10. [10]

    Jiayang Chen, Zhihang Hu, Siqi Sun, Qingxiong Tan, {WANG Yixuan}, Qinze Yu, Licheng Zong, Liang Hong, Jin Xiao, Irwin King, and {LI Yu}. 2022. Interpretable RNA foundation model from unannotated data for highly accurate RNA structure and function predictions. The 2022 ICML Workshop on Computational Biology ; Conference date: 17-07-2022 Through 23-07-2022

    2022

  11. [11]

    Hugo Dalla-Torre, Liam Gonzalez, Javier Mendoza-Revilla, Nicolas Lopez Car- ranza, Adam Henryk Grzywaczewski, Francesco Oteri, Christian Dallago, Evan Trop, Bernardo P de Almeida, Hassan Sirelkhatim, et al. 2025. Nucleotide Trans- former: building and evaluating robust foundation models for human genomics. Nature Methods22, 2 (2025), 287–297

    2025

  12. [12]

    Christian Dallago, Jody Mou, Kadina E Johnston, Bruce Wittmann, Nick Bhat- tacharya, Samuel Goldman, Ali Madani, and Kevin K Yang. 2021. FLIP: Benchmark tasks in fitness landscape inference for proteins. InThirty-fifth Conference on Neural Information Processing Systems Datasets and Benchmarks Track (Round 2). https://openreview.net/forum?id=p2dMLEwL8tF

    2021

  13. [13]

    Xing, and Le Song

    Caleb Ellington, Ning Sun, Nicholas Ho, Tianhua Tao, Sazan Mahbub, Dian Li, Yonghao Zhuang, Hongyi Wang, Eric P. Xing, and Le Song. 2024. Accurate and General DNA Representations Emerge from Genome Foundation Models at Scale. InNeurIPS 2024 Workshop on AI for New Drug Modalities. https://openreview. net/forum?id=Kis8tVUeNi

    2024

  14. [14]

    Alfred Ferrer Florensa, Jose Juan Almagro Armenteros, Henrik Nielsen, Frank Møller Aarestrup, and Philip Thomas Lanken Conradsen Clausen. 2024. SpanSeq: similarity-based sequence data splitting method for improved develop- ment and assessment of deep learning projects.NAR Genomics and Bioinformatics 6, 3 (2024), lqae106

    2024

  15. [15]

    Veniamin Fishman, Yuri Kuratov, Aleksei Shmelev, Maxim Petrov, Dmitry Penzar, Denis Shepelin, Nikolay Chekanov, Olga Kardymon, and Mikhail Burtsev. 2025. GENA-LM: a family of open-source foundational DNA language models for long sequences.Nucleic Acids Research53, 2 (2025), gkae1310

    2025

  16. [16]

    Zhangyang Gao, Cheng Tan, Yijie Zhang, Xingran Chen, Lirong Wu, and Stan Z Li. 2023. Proteininvbench: Benchmarking protein inverse folding on diverse tasks, models, and metrics.Advances in Neural Information Processing Systems36 (2023), 68207–68220

    2023

  17. [17]

    Cody Glickman, Jo Hendrix, and Michael Strong. 2021. Simulation study and comparative evaluation of viral contiguous sequence identification tools.BMC bioinformatics22, 1 (2021), 329

    2021

  18. [18]

    Katarína Grešová, Vlastimil Martinek, David Čechák, Petr Šimeček, and Panagi- otis Alexiou. 2023. Genomic benchmarks: a collection of datasets for genomic sequence classification.BMC Genomic Data24, 1 (2023), 25

    2023

  19. [19]

    Yong He, Pan Fang, Yongtao Shan, Yuanfei Pan, Yanhong Wei, Yichang Chen, Yihao Chen, Yi Liu, Zhenyu Zeng, Zhan Zhou, et al. 2025. Generalized biological foundation model with unified nucleic acid and protein language.Nature Machine Intelligence(2025), 1–12

    2025

  20. [20]

    Edward C Holmes. 2013. What can we predict about viral evolution and emer- gence?Current opinion in virology3, 2 (2013), 180–184

    2013

  21. [21]

    Yanrong Ji, Zhihan Zhou, Han Liu, and Ramana V Davuluri. 2021. DNABERT: pre-trained Bidirectional Encoder Representations from Transformers model for DNA-language in genome.Bioinformatics37, 15 (2021), 2112–2120

    2021

  22. [22]

    Zhongmin Li, Runze Ma, Jiahao Tan, Chengzi Tan, and Shuangjia Zheng. 2025. NABench: Large-Scale Benchmarks of Nucleotide Foundation Models for Fitness Prediction.arXiv preprint arXiv:2511.02888(2025)

    work page arXiv 2025

  23. [23]

    Adi Lin, Bin Xie, Cheng Ye, Cheng Wang, Duoyuan Chen, Ercheng Wang, Fanfeng Lu, Guirong Xue, Haiqiang Zhang, Jiajie Zhan, et al. 2025. Genos: a human-centric genomic foundation model.GigaScience14 (2025), giaf132

    2025

  24. [24]

    LeAnn M Lindsey, Nicole L Pershing, Anisa Habib, Keith Dufault-Thompson, W Zac Stephens, Anne J Blaschke, Xiaofang Jiang, and Hari Sundar. 2025. The impact of tokenizer selection in genomic language models.Bioinformatics41, 9 (2025), btaf456

    2025

  25. [25]

    Zicheng Liu, Jiahui Li, Siyuan Li, Zelin Zang, Cheng Tan, Yufei Huang, Yajing Bai, and Stan Z Li. 2024. Genbench: A benchmarking suite for systematic evaluation of genomic foundation models.arXiv preprint arXiv:2406.01627(2024)

    work page arXiv 2024

  26. [26]

    Frederikke Isa Marin, Felix Teufel, Marc Horlacher, Dennis Madsen, Dennis Pultz, Ole Winther, and Wouter Boomsma. 2024. BEND: Benchmarking DNA Language Models on Biologically Meaningful Tasks. InThe Twelfth International Conference on Learning Representations. https://openreview.net/forum?id=uKB4cFNQFg

    2024

  27. [27]

    Aditi T Merchant, Samuel H King, Eric Nguyen, and Brian L Hie. 2026. Semantic design of functional de novo genes from a genomic language model.Nature649, 8097 (2026), 749–758

    2026

  28. [28]

    Hayden C Metsky, Nicole L Welch, Priya P Pillai, Nicholas J Haradhvala, Laurie Rumker, Sreekar Mantena, Yibin B Zhang, David K Yang, Cheri M Ackerman, Juliane Weller, et al. 2022. Designing sensitive viral diagnostics with machine learning.Nature biotechnology40, 7 (2022), 1123–1131

    2022

  29. [29]

    Florian Mock, Adrian Viehweger, Emanuel Barth, and Manja Marz. 2021. VIDHOP, viral host prediction with deep learning.Bioinformatics37, 3 (2021), 318–325

    2021

  30. [30]

    Eric Nguyen, Michael Poli, Matthew G Durrant, Brian Kang, Dhruva Katrekar, David B Li, Liam J Bartie, Armin W Thomas, Samuel H King, Garyk Brixi, et al

  31. [31]

    Science386, 6723 (2024), eado9336

    Sequence modeling and design from molecular to genome scale with Evo. Science386, 6723 (2024), eado9336

    2024

  32. [32]

    Eric Nguyen, Michael Poli, Marjan Faizi, Armin Thomas, Michael Wornow, Cal- lum Birch-Sykes, Stefano Massaroli, Aman Patel, Clayton Rabideau, Yoshua Bengio, et al. 2023. Hyenadna: Long-range genomic sequence modeling at single nucleotide resolution.Advances in neural information processing systems36 (2023), 43177–43201

    2023

  33. [33]

    Pascal Notin, Aaron Kollasch, Daniel Ritter, Lood Van Niekerk, Steffanie Paul, Han Spinner, Nathan Rollins, Ada Shaw, Rose Orenbuch, Ruben Weitzman, et al

  34. [34]

    Proteingym: Large-scale benchmarks for protein fitness prediction and design.Advances in neural information processing systems36 (2023), 64331–64379

    2023

  35. [35]

    Nuala A O’Leary, Mathew W Wright, J Rodney Brister, Stacy Ciufo, Diana Haddad, Rich McVeigh, Bhanu Rajput, Barbara Robbertse, Brian Smith-White, Danso Ako- Adjei, et al. 2016. Reference sequence (RefSeq) database at NCBI: current status, taxonomic expansion, and functional annotation.Nucleic acids research44, D1 (2016), D733–D745

    2016

  36. [36]

    Yuan-Fei Pan, Yong He, Yu-Qi Liu, Yong-Tao Shan, Shu-Ning Liu, Jia-Hao Ma, Xue Liu, Xiaoyun Pan, Yinqi Bai, Zan Xu, et al. 2025. Predicting the evolutionary and functional landscapes of viruses with a unified nucleotide-protein language model: Lucavirus.bioRxiv(2025), 2025–06

    2025

  37. [37]

    Rafael Josip Penić, Tin Vlašić, Roland G Huber, Yue Wan, and Mile Šikić. 2025. Rinalmo: General-purpose rna language models can generalize well on structure prediction tasks.Nature Communications16, 1 (2025), 5671

    2025

  38. [38]

    Fedor S Perelygin, Alexander N Lukashev, and Yulia A Aleshina. 2025. The effect of taxonomic, host-dependent features and sample bias on virus host prediction using machine learning and short sequence k-mers.Scientific reports15, 1 (2025), 31592

    2025

  39. [39]

    Purwono Purwono, Annastasya Nabila Elsa Wulandari, and Novieta Hardeani Sari. 2024. Virus Host Prediction with Metagenomic Features using Support Vec- tor Machine Algorithm and Grid Search Cross Validation Optimization.Journal of Advanced Health Informatics Research2, 3 (2024), 127–137

    2024

  40. [40]

    Rajan Saha Raju, Abdullah Al Nahid, Preonath Chondrow Dev, and Rashedul Islam. 2022. VirusTaxo: Taxonomic classification of viruses from the genome sequence using k-mer enrichment.Genomics114, 4 (2022), 110414

    2022

  41. [41]

    Jie Ren, Nathan A Ahlgren, Yang Young Lu, Jed A Fuhrman, and Fengzhu Sun

  42. [42]

    VirFinder: a novel k-mer based tool for identifying viral sequences from assembled metagenomic data.Microbiome5, 1 (2017), 69

    2017

  43. [43]

    Jie Ren, Kai Song, Chao Deng, Nathan A Ahlgren, Jed A Fuhrman, Yi Li, Xiaohui Xie, Ryan Poplin, and Fengzhu Sun. 2020. Identifying viruses from metagenomic data using deep learning.Quantitative Biology8, 1 (2020), 64–77

    2020

  44. [44]

    Melissa Sanabria, Jonas Hirsch, Pierre M Joubert, and Anna R Poetsch. 2024. DNA language model GROVER learns sequence context in the human genome.Nature Machine Intelligence6, 8 (2024), 911–923

    2024

  45. [45]

    Rafael Sanjuán, Miguel R Nebot, Nicola Chirico, Louis M Mansky, and Robert Belshaw. 2010. Viral mutation rates.Journal of virology84, 19 (2010), 9733–9748

    2010

  46. [46]

    Josep Sardanyés, Celia Perales, Esteban Domingo, and Santiago F Elena. 2024. Quasispecies theory and emerging viruses: challenges and applications.npj Viruses2, 1 (2024), 54. ViroBench: Benchmarking Nucleotide Foundation Models on Viral Genomics Tasks KDD ’26, August 9–13, 2026, Jeju,Korea

    2024

  47. [47]

    Yair Schiff, Chia Hsiang Kao, Aaron Gokaslan, Tri Dao, Albert Gu, and Volodymyr Kuleshov. 2024. Caduceus: Bi-Directional Equivariant Long-Range DNA Sequence Modeling. InProceedings of the 41st International Conference on Machine Learning (Proceedings of Machine Learning Research, Vol. 235), Ruslan Salakhutdinov, Zico Kolter, Katherine Heller, Adrian Welle...

    2024

  48. [48]

    Mike Schuster and Kuldip K Paliwal. 1997. Bidirectional recurrent neural net- works.IEEE transactions on Signal Processing45, 11 (1997), 2673–2681

    1997

  49. [49]

    Jiayu Shang and Yanni Sun. 2021. CHEER: HierarCHical taxonomic classification for viral mEtagEnomic data via deep leaRning.Methods189 (2021), 95–103

    2021

  50. [50]

    Md Toki Tahmid, Haz Sameen Shahgir, Sazan Mahbub, Yue Dong, and Md Sham- suzzoha Bayzid. 2025. BiRNA-BERT allows efficient RNA language modeling with adaptive tokenization.Communications Biology8, 1 (2025), 1621

    2025

  51. [51]

    Aowen Wang, Jiaqi Li, Hongyu Dong, Bocheng Xu, Qingyu Yin, Yanchao Xu, Jie Fu, and Junbo Zhao. 2025. Omnireg-gpt: a high-efficiency foundation model for comprehensive genomic sequence understanding.Nature Communications16, 1 (2025), 10139

    2025

  52. [52]

    James D Watson and Francis HC Crick. 1953. Molecular structure of nucleic acids: a structure for deoxyribose nucleic acid.Nature171, 4356 (1953), 737–738

    1953

  53. [53]

    Nicole E Wheeler. 2025. Responsible AI in biotechnology: balancing discovery, innovation and biosecurity risks.Frontiers in Bioengineering and Biotechnology 13 (2025), 1537471

    2025

  54. [54]

    Derrick E Wood, Jennifer Lu, and Ben Langmead. 2019. Improved metagenomic analysis with Kraken 2.Genome biology20, 1 (2019), 257

    2019

  55. [55]

    Wei Wu, Qiuyi Li, Yuanyuan Zhang, Zhihao Zhan, Ruipu Chen, Mingyang Li, Kun Fu, Junyan Qi, Yongzhou Bao, Chao Wang, et al. 2025. GENERator: a long-context generative genomic foundation model.arXiv preprint arXiv:2502.07272(2025)

    work page arXiv 2025

  56. [56]

    Minghao Xu, Zuobai Zhang, Jiarui Lu, Zhaocheng Zhu, Yangtian Zhang, Ma Chang, Runcheng Liu, and Jian Tang. 2022. Peer: a comprehensive and multi-task benchmark for protein sequence understanding.Advances in Neural Information Processing Systems35 (2022), 35156–35173

    2022

  57. [57]

    An Yang, Anfeng Li, Baosong Yang, Beichen Zhang, Binyuan Hui, Bo Zheng, Bowen Yu, Chang Gao, Chengen Huang, Chenxu Lv, et al. 2025. Qwen3 technical report.arXiv preprint arXiv:2505.09388(2025)

    work page internal anchor Pith review Pith/arXiv arXiv 2025

  58. [58]

    Heng Yang and Ke Li. 2024. MP-RNA: Unleashing Multi-species RNA Foun- dation Model via Calibrated Secondary Structure Prediction. InFindings of the Association for Computational Linguistics: EMNLP 2024. 5278–5296

    2024

  59. [59]

    Francisco Murilo Zerbini, Stuart G Siddell, Elliot J Lefkowitz, Arcady R Mushegian, Evelien M Adriaenssens, Poliane Alfenas-Zerbini, Donald M Dempsey, Bas E Dutilh, Maria Laura Garcia, R Curtis Hendrickson, et al. 2023. Changes to virus taxonomy and the ICTV Statutes ratified by the International Committee on Taxonomy of Viruses (2023).Archives of virolog...

    2023

  60. [60]

    Liang Zhang, Hua Pang, Chenghao Zhang, Song Li, Yang Tan, Fan Jiang, Mingchen Li, Yuanxi Yu, Ziyi Zhou, Banghao Wu, et al . 2025. VenusMutHub: a systematic evaluation of protein mutation effect predictors on small-scale experimental data.Acta Pharmaceutica Sinica B(2025)

    2025

  61. [61]

    Zhihan Zhou, Yanrong Ji, Weijian Li, Pratik Dutta, Ramana Davuluri, and Han Liu

  62. [62]

    InInternational Conference on Learning Representations, Vol

    DNABERT-2: Efficient foundation model and benchmark for multi-species genomes. InInternational Conference on Learning Representations, Vol. 2024. 41642–41665

    2024

  63. [63]

    Zhihan Zhou, Robert Riley, Satria Kautsar, Weimin Wu, Rob Egan, Steven Hofmeyr, Shira Goldhaber-Gordon, Mutian Yu, Harrison Ho, Fengchen Liu, et al

  64. [64]

    GenomeOcean: An Efficient Genome Foundation Model Trained on Large- Scale Metagenomic Assemblies.bioRxiv(2025), 2025–01

    2025

  65. [65]

    Zhihan Zhou, Weimin Wu, Harrison Ho, Jiayi Wang, Lizhen Shi, Ramana V Davuluri, Zhong Wang, and Han Liu. 2025. DNABERT-S: Pioneering species differentiation with species-aware DNA embeddings.Bioinformatics41, Supple- ment_1 (2025), i255–i264

    2025

  66. [66]

    Shuxian Zou, Tianhua Tao, Sazan Mahbub, Caleb Ellington, Robin Jonathan Algayres, Dian Li, Yonghao Zhuang, Hongyi Wang, Le Song, and Eric P. Xing

  67. [67]

    recorded time

    A Large-Scale Foundation Model for RNA Function and Structure Prediction. InNeurIPS 2024 Workshop on AI for New Drug Modalities. https://openreview. net/forum?id=Gzo3JMPY8w Appendix Table of Contents A Detailed Data Curation Pipeline 12 A.1 Data Sources and Quality Filters 12 A.2 Data Partitioning 12 A.3 Recommended Lightweight Evaluation Subset 13 B Impl...

    2024