pocketz.doi.bio/2026-01-11_200008_nature_s41401-025-01688-3

Graph-based deep learning approach for high-throughput protein-DNA interaction scoring | Acta Pharmacologica Sinica

URL: https://www.nature.com/articles/s41401-025-01688-3

Saved: 2026-01-11T04:00:08.929Z

Article Published: 01 December 2025 Graph-based deep learning approach for high-throughput protein-DNA interaction scoring Yi-hao Zhao, Ying Wang, Chao Shen, De-jun Jiang, Shu-kai Gu, Hui-feng Zhao, Zi-yi You, Ting-jun Hou & Yu Kang 

Acta Pharmacologica Sinica (2025)Cite this article

365 Accesses

1 Altmetric

Metrics details

Abstract

Accurately quantifying protein-DNA interactions (PDIs) is critical for understanding biological processes and facilitating drug design. However, the inherent flexibility of nucleic acids limits the availability of experimentally determined structures of PDI complexes, posing a significant challenge for training reliable scoring functions (SFs). To address this, we developed PDIScore, a novel deep learning-based SF for PDI prediction. PDIScore utilizes a comprehensive graph representation to capture nucleotide flexibility, employs a scalable GraphGPS architecture with BigBird linear global attention to handle large interaction interfaces, and leverages Mixture Density Networks (MDNs) to model residue-nucleotide distance distributions. PDIScore was trained on a self-collected dataset of ~7000 protein-nucleic acid complex structures and validated on three rigorous test sets for evaluating its screening, docking, and ranking capabilities. The results illustrated that PDIScore significantly outperformed existing methods: it achieved the best screening power on the screening set (e.g., EF1% = 14.13, AUROC = 0.82 using AlphaFold3 structures), the highest docking success rate on the docking set (48.94% top1), and superior ranking capability on the ranking set (PCC = 0.50). Case studies demonstrated PDIScore’s ability to elucidate biological mechanisms (e.g., adenovirus transcription, SOCS1 regulation) and its interpretability at the nucleotide level for identifying key interaction sites. PDIScore represents a robust, generalizable tool with significant potential for advancing PDI-related research and therapeutic design.

This is a preview of subscription content, access via your institution

Access options

Access to this article via is not available.

Change institution

Subscribe to this journal

Receive 12 print issues and online access

$259.00 per year

only $21.58 per issue

Learn more

Buy this article

Purchase on SpringerLink Instant access to the full article PDF.

USD 39.95

Prices may be subject to local taxes which are calculated during checkout

Additional access options: Log in Learn about institutional subscriptions Read our FAQs Contact customer support Similar content being viewed by others Predicting Parkinson disease related genes based on PyFeat and gradient boosted decision tree Article Open access 15 June 2022 Decoding the protein–ligand interactions using parallel graph neural networks Article Open access 10 May 2022 A robust deep learning approach for identification of RNA 5-methyluridine sites Article Open access 28 October 2024 Data availability

The datasets are available at https://doi.org/10.5281/zenodo.13764615. The code is available at https://github.com/roger-yh-zhao/pdiscore.

References

Charoensawan V, Wilson D, Teichmann SA. Genomic repertoires of DNA-binding transcription factors across the tree of life. Nucleic Acids Res. 2010;38:7364–77.

Article   PubMed   PubMed Central   CAS   Google Scholar

Markodimitraki CM, Rang FJ, Rooijers K, de Vries SS, Chialastri A, de Luca KL, et al. Simultaneous quantification of protein-DNA interactions and transcriptomes in single cells with scDam&T-seq. Nat Protoc. 2020;15:1922–53.

Article   PubMed   PubMed Central   CAS   Google Scholar

Yu L, Liu P. Cytosolic DNA sensing by cGAS: regulation, function, and human diseases. Signal Transduct Target Ther. 2021;6:170.

Article   PubMed   PubMed Central   CAS   Google Scholar

Wells A, Heckerman D, Torkamani A, Yin L, Sebat J, Ren B, et al. Ranking of non-coding pathogenic variants and putative essential regions of the human genome. Nat Commun. 2019;10:5241.

Article   PubMed   PubMed Central   Google Scholar

Jiao W, Atwal G, Polak P, Karlic R, Cuppen E, Subtypes PT, et al. A deep learning system accurately classifies primary and metastatic cancers using passenger mutation patterns. Nat Commun. 2020;11:728.

Article   PubMed   PubMed Central   CAS   Google Scholar

Guo AD, Yan KN, Hu H, Zhai L, Hu TF, Su H, et al. Spatiotemporal and global profiling of DNA-protein interactions enables discovery of low-affinity transcription factors. Nat Chem. 2023;15:803–14.

Article   PubMed   CAS   Google Scholar

Chan LL, Pineda M, Heeres JT, Hergenrother PJ, Cunningham BT. A general method for discovering inhibitors of protein-DNA interactions using photonic crystal biosensors. ACS Chem Biol. 2008;3:437–48.

Article   PubMed   PubMed Central   CAS   Google Scholar

Barabasi AL, Gulbahce N, Loscalzo J. Network medicine: a network-based approach to human disease. Nat Rev Genet. 2011;12:56–68.

Article   PubMed   PubMed Central   CAS   Google Scholar

Nikolov DB, Chen H, Halay ED, Usheva AA, Hisatake K, Lee DK, et al. Crystal structure of a TFIIB-TBP-TATA-element ternary complex. Nature. 1995;377:119–28.

Article   PubMed   CAS   Google Scholar

Hashimoto H, Olanrewaju YO, Zheng Y, Wilson GG, Zhang X, Cheng X. Wilms tumor protein recognizes 5-carboxylcytosine within a specific DNA sequence. Genes Dev. 2014;28:2304–13.

Article   PubMed   PubMed Central   Google Scholar

Kays AR, Schepartz A. Virtually unidirectional binding of TBP to the AdMLP TATA box within the quaternary complex with TFIIA and TFIIB. Chem Biol. 2000;7:601–10.

Article   PubMed   CAS   Google Scholar

Mostecki J, Showalter BM, Rothman PB. Early growth response-1 regulates lipopolysaccharide-induced suppressor of cytokine signaling-1 transcription. J Biol Chem. 2005;280:2596–605.

Article   PubMed   CAS   Google Scholar

Sobah ML, Liongue C, Ward AC. SOCS proteins in immunity, inflammatory diseases, and immune-related cancer. Front Med. 2021;8:727987.

Article   Google Scholar

Sefah K, Shangguan D, Xiong X, O’Donoghue MB, Tan W. Development of DNA aptamers using Cell-SELEX. Nat Protoc. 2010;5:1169–85.

Article   PubMed   CAS   Google Scholar

Huynh L, Chen A. Design of a protein-targeted DNA aptamer using atomistic simulation. J Biomol Struct Dyn. 2023;41:672–80.

Article   PubMed   CAS   Google Scholar

Wang X, Wang Y, Cao A, Luo Q, Chen D, Zhao W, et al. Development of cyclopeptide inhibitors of cGAS targeting protein-DNA interaction and phase separation. Nat Commun. 2023;14:6132.

Article   PubMed   PubMed Central   CAS   Google Scholar

Kimoto M, Yamashige R, Matsunaga K, Yokoyama S, Hirao I. Generation of high-affinity DNA aptamers using an expanded genetic alphabet. Nat Biotechnol. 2013;31:453–7.

Article   PubMed   CAS   Google Scholar

Hellman LM, Fried MG. Electrophoretic mobility shift assay (EMSA) for detecting protein-nucleic acid interactions. Nat Protoc. 2007;2:1849–61.

Article   PubMed   PubMed Central   CAS   Google Scholar

Hadzi S, Lah J. Analysis of protein-DNA interactions using isothermal titration calorimetry: successes and failures. Methods Mol Biol. 2022;2516:239–57.

Article   PubMed   CAS   Google Scholar

Majka J, Speck C. Analysis of protein-DNA interactions using surface plasmon resonance. Adv Biochem Eng Biotechnol. 2007;104:13–36.

PubMed   CAS   Google Scholar

Townshend RJL, Eismann S, Watkins AM, Rangan R, Karelina M, Das R, et al. Geometric deep learning of RNA structure. Science. 2021;373:1047–51.

Article   PubMed   PubMed Central   CAS   Google Scholar

Roche R, Moussad B, Shuvo MH, Tarafder S, Bhattacharya D. EquiPNAS: improved protein-nucleic acid binding site prediction using protein-language-model-informed equivariant deep graph neural networks. Nucleic Acids Res. 2024;52:e27.

Article   PubMed   PubMed Central   CAS   Google Scholar

Delgado J, Radusky LG, Cianferoni D, Serrano L. FoldX 5.0: working with RNA, small molecules and a new graphical interface. Bioinformatics. 2019;35:4168–9.

Article   PubMed   PubMed Central   CAS   Google Scholar

Schymkowitz J, Borg J, Stricher F, Nys R, Rousseau F, Serrano L. The FoldX web server: an online force field. Nucleic Acids Res. 2005;33:W382–8.

Article   PubMed   PubMed Central   CAS   Google Scholar

Wang E, Sun H, Wang J, Wang Z, Liu H, Zhang JZH, et al. End-point binding free energy calculation with MM/PBSA and MM/GBSA: strategies and applications in drug design. Chem Rev. 2019;119:9478–508.

Article   PubMed   CAS   Google Scholar

Chaudhury S, Lyskov S, Gray JJ. PyRosetta: a script-based interface for implementing molecular modeling algorithms using Rosetta. Bioinformatics. 2010;26:689–91.

Article   PubMed   PubMed Central   CAS   Google Scholar

Yang W, Deng L. PreDBA: a heterogeneous ensemble approach for predicting protein-DNA binding affinity. Sci Rep. 2020;10:1278.

Article   PubMed   PubMed Central   CAS   Google Scholar

Yang S, Gong W, Zhou T, Sun X, Chen L, Zhou W, et al. emPDBA: protein-DNA binding affinity prediction by combining features from binding partners and interface learned with ensemble regression model. Brief Bioinform. 2023;24:bbad192.

Zhang X, Mei LC, Gao YY, Hao GF, Song BA. Web tools support predicting protein-nucleic acid complexes stability with affinity changes. WIREs RNA. 2023;14:e1781.

Article   PubMed   Google Scholar

Yan Y, Zhang D, Zhou P, Li B, Huang SY. HDOCK: a web server for protein-protein and protein-DNA/RNA docking based on a hybrid strategy. Nucleic Acids Res. 2017;45:W365–73.

Article   PubMed   PubMed Central   CAS   Google Scholar

Rodriguez-Lumbreras LA, Jimenez-Garcia B, Gimenez-Santamarina S, Fernandez-Recio J. pyDockDNA: a new web server for energy-based protein-DNA docking and scoring. Front Mol Biosci. 2022;9:988996.

Article   PubMed   PubMed Central   CAS   Google Scholar

van Zundert GCP, Rodrigues J, Trellet M, Schmitz C, Kastritis PL, Karaca E, et al. The HADDOCK2.2 web server: user-friendly integrative modeling of biomolecular complexes. J Mol Biol. 2016;428:720–5.

Article   PubMed   Google Scholar

Honorato RV, Koukos PI, Jimenez-Garcia B, Tsaregorodtsev A, Verlato M, Giachetti A, et al. Structural biology in the clouds: the WeNMR-EOSC ecosystem. Front Mol Biosci. 2021;8:729513.

Article   PubMed   PubMed Central   Google Scholar

Jumper J, Evans R, Pritzel A, Green T, Figurnov M, Ronneberger O, et al. Highly accurate protein structure prediction with AlphaFold. Nature. 2021;596:583–9.

Article   PubMed   PubMed Central   CAS   Google Scholar

Abramson J, Adler J, Dunger J, Evans R, Green T, Pritzel A, et al. Accurate structure prediction of biomolecular interactions with AlphaFold 3. Nature. 2024;630:493–500.

Article   PubMed   PubMed Central   CAS   Google Scholar

Berman HM, Westbrook J, Feng Z, Gilliland G, Bhat TN, Weissig H, et al. The Protein Data Bank. Nucleic Acids Res. 2000;28:235–42.

Article   PubMed   PubMed Central   CAS   Google Scholar

Afek A, Shi H, Rangadurai A, Sahay H, Senitzki A, Xhani S, et al. DNA mismatches reveal conformational penalties in protein-DNA recognition. Nature. 2020;587:291–6.

Article   PubMed   PubMed Central   CAS   Google Scholar

van Dijk M, Bonvin AM. A protein-DNA docking benchmark. Nucleic Acids Res. 2008;36:e88.

Article   PubMed   PubMed Central   Google Scholar

Larkin C, Datta S, Harley MJ, Anderson BJ, Ebie A, Hargreaves V, et al. Inter- and intramolecular determinants of the specificity of single-stranded DNA binding and cleavage by the F factor relaxase. Structure. 2005;13:1533–44.

Article   PubMed   CAS   Google Scholar

Li J, Lee JC. Modulation of allosteric behavior through adjustment of the differential stability of the two interacting domains in E. coli cAMP receptor protein. Biophys Chem. 2011;159:210–6.

Article   PubMed   PubMed Central   CAS   Google Scholar

Yu S, Maillard RA, Gribenko AV, Lee JC. The N-terminal capping propensities of the D-helix modulate the allosteric activation of the Escherichia coli cAMP receptor protein. J Biol Chem. 2012;287:39402–11.

Article   PubMed   PubMed Central   CAS   Google Scholar

Seldeen KL, Deegan BJ, Bhat V, Mikles DC, McDonald CB, Farooq A. Energetic coupling along an allosteric communication channel drives the binding of Jun-Fos heterodimeric transcription factor to DNA. FEBS J. 2011;278:2090–104.

Article   PubMed   PubMed Central   CAS   Google Scholar

Rampášek L, Galkin M, Dwivedi VP, Luu AT, Wolf G, Beaini D. Recipe for a general, powerful, scalable graph transformer. NeurIPS. 2022;35:14501–15.

Google Scholar

Zaheer M, Guruganesh G, Dubey KA, Ainslie J, Alberti C, Ontanon S, et al. Big bird: transformers for longer sequences. NeurIPS. 2020;33:17283–97.

Google Scholar

Shen C, Zhang X, Hsieh CY, Deng Y, Wang D, Xu L, et al. A generalized protein-ligand scoring framework with balanced scoring, docking, ranking and screening powers. Chem Sci. 2023;14:8129–46.

Article   PubMed   PubMed Central   CAS   Google Scholar

Shen C, Zhang X, Deng Y, Gao J, Wang D, Xu L, et al. Boosting protein-ligand binding pose prediction and virtual screening based on residue-atom distance likelihood potential and graph transformer. J Med Chem. 2022;65:10691–706.

Article   PubMed   CAS   Google Scholar

Wang R, Fang X, Lu Y, Wang S. The PDBbind database: collection of binding affinities for protein-ligand complexes with known three-dimensional structures. J Med Chem. 2004;47:2977–80.

Article   PubMed   CAS   Google Scholar

Pettersen EF, Goddard TD, Huang CC, Couch GS, Greenblatt DM, Meng EC, et al. UCSF Chimera-a visualization system for exploratory research and analysis. J Comput Chem. 2004;25:1605–12.

Article   PubMed   CAS   Google Scholar

Leaver-Fay A, Tyka M, Lewis SM, Lange OF, Thompson J, Jacak R, et al. ROSETTA3: an object-oriented software suite for the simulation and design of macromolecules. Methods Enzymol. 2011;487:545–74.

Article   PubMed   PubMed Central   CAS   Google Scholar

Lu XJ, Olson WK. 3DNA: a software package for the analysis, rebuilding and visualization of three-dimensional nucleic acid structures. Nucleic Acids Res. 2003;31:5108–21.

Article   PubMed   PubMed Central   CAS   Google Scholar

Steinegger M, Soding J. MMseqs2 enables sensitive protein sequence searching for the analysis of massive data sets. Nat Biotechnol. 2017;35:1026–8.

Article   PubMed   CAS   Google Scholar

Michaud-Agrawal N, Denning EJ, Woolf TB, Beckstein O. MDAnalysis: a toolkit for the analysis of molecular dynamics simulations. J Comput Chem. 2011;32:2319–27.

Article   PubMed   PubMed Central   CAS   Google Scholar

Wang M, Xu L, Zheng D, Gan Q, Gai Y, Ye Z, et al. Deep graph library: towards efficient and scalable deep learning on graphs. Preprint at. https://doi.org/10.48550/arXiv.1909.01315 (2019).

Hu W, Liu B, Gomes J, Zitnik M, Liang P, Pande V, et al. In presented in part at the International Conference on Learning Representations (ICLR), (ICLR, 2020).

Bauer MR, Ibrahim TM, Vogel SM, Boeckler FM. Evaluation and optimization of virtual screening workflows with DEKOIS 2.0-a public library of challenging docking benchmark sets. J Chem Inf Model. 2013;53:1447–62.

Article   PubMed   CAS   Google Scholar

Concino M, Goldman RA, Caruthers MH, Weinmann R. Point mutations of the adenovirus major late promoter with different transcriptional efficiencies in vitro. J Biol Chem. 1983;258:8493–6.

Article   PubMed   CAS   Google Scholar

Ilangumaran S, Gui Y, Shukla A, Ramanathan S. SOCS1 expression in cancer cells: potential roles in promoting antitumor immunity. Front Immunol. 2024;15:1362224.

Article   PubMed   PubMed Central   CAS   Google Scholar

Download references

Acknowledgements

This work was financially supported by National Natural Science Foundation of China (22220102001, 22307112, 92370130, 22303081) and Postdoctoral Fellowship Program of CPSF (GZB20230648).

Author information Author notes

These authors contributed equally: Yi-hao Zhao, Ying Wang.

Authors and Affiliations

College of Pharmaceutical Sciences, Zhejiang University, Hangzhou, 310058, China

Yi-hao Zhao, Ying Wang, Shu-kai Gu, Hui-feng Zhao, Zi-yi You, Ting-jun Hou & Yu Kang

Department of Clinical Pharmacy, the First Affiliated Hospital of Zhejiang University School of Medicine, Hangzhou, 310003, China

Chao Shen

Xiangya School of Pharmaceutical Sciences, Central South University, Changsha, 410004, China

De-jun Jiang

Zhejiang Provincial Key Laboratory for Intelligent Drug Discovery and Development, Jinhua, 321016, China

Ting-jun Hou & Yu Kang

Contributions

TJH and YK designed the research study. YHZ developed the method and wrote the code. YHZ, YW, CS, DJJ, SKG, HFZ, and ZYY performed the analysis. YHZ and TJH wrote the paper. All authors read and approved the manuscript.

Corresponding authors

Correspondence to Ting-jun Hou or Yu Kang.

Ethics declarations Competing interests

The authors declare no competing interests.

Additional information

Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Supplementary information Supplementary information affinity cluster Rights and permissions

Springer Nature or its licensor (e.g. a society or other partner) holds exclusive rights to this article under a publishing agreement with the author(s) or other rightsholder(s); author self-archiving of the accepted manuscript version of this article is solely governed by the terms of such publishing agreement and applicable law.

Reprints and permissions

About this article Cite this article

Zhao, Yh., Wang, Y., Shen, C. et al. Graph-based deep learning approach for high-throughput protein-DNA interaction scoring. Acta Pharmacol Sin (2025). https://doi.org/10.1038/s41401-025-01688-3

Download citation

Received 14 July 2025

Accepted 02 October 2025

Published 01 December 2025

Version of record 01 December 2025

DOI https://doi.org/10.1038/s41401-025-01688-3

Key words protein-DNA interactions machine learning deep learning molecular docking virtual screening