DeepETD: a novel deep-learning based model for endogenous metabolite target discovery

Abstract

Metabolites are involved in almost all fundamental biological processes. The identification of their protein targets is crucial for elucidating non-canonical signaling roles and evaluating their therapeutic potential. In recent years, due to the continuous development of chemical proteomics, atypical phenotypic functions of classic metabolites have been continuously discovered. However, the discovery of endogenous metabolites for their disease-related functions is still progressing slowly. To accelerate the identification of disease-related targets for endogenous metabolites, we propose a new hypothesis: endogenous metabolites and their molecular targets are expected to have similar disease associations. Following this hypothesis, here we report the development of a novel deep-learning model called DeepETD, which integrates bioinformatics data and introduces an attention mechanism to predict functional targets of specific metabolite phenotypes. Using this model, we constructed a publicly accessible database named EMTDD containing potential targets for 3,382 common human endogenous metabolites. Overall, this study presents a new computational method and resource for endogenous metabolite target discovery as an important supplement to experimental methods such as chemical proteomics.

Ethical statements

Not applicable. The present study included only in vitro experiments and did not involve human subjects, clinical samples, or animal experiments.

Author contributions

The authors confirm their contributions to the work as follows: development of the computational tool and the writing of the corresponding methodological and related content: Yang X, Xu Z, Min X; manuscript writing: Xu Z, Wang X; experimental validation and figure preparation: Wang X, Yuan X; reference collection and organization: Xu Z, Zhu K, Xiao W; manuscript checking and revision: Zhang H, Xu H, Luo C. All authors reviewed the results and approved the final version of the manuscript.

Data availability

The code of DeepETD is available at https://github.com/AIDDHao/DeepETD. The data generated or analyzed during this study are included in this published article and its supplementary information files. Additional data related to this study are available from the corresponding author upon reasonable request.

Acknowledgment

We thank the staff members of the Large-scale Protein Preparation System at the National Facility for Protein Science in Shanghai, for providing technical support and assistance in data collection and analysis. We gratefully acknowledge the financial supports from the National Key R&D Program of China (2022YFC3400500 to Cheng Luo), the Strategic Priority Research Program of the Chinese Academy of Sciences (XDB0830301 to Cheng Luo), the National Natural Science Foundation of China (81903538 to Hao Zhang), the Science and Technology Department of Guizhou Province (CXPTXM[2025]021 and KXJZ[2025]014 to Cheng Luo), the Applied Basic Research Foundation of Yunnan Province (202501BC070005 to Cheng Luo), the Shanghai Municipal Health Commission Medical New Technology Project (2025ZZ2060 to Hao Zhang), the Shanghai Oriental Talent Plan Youth Project (QNJY2025170 to Hao Zhang), the Shanghai Science and Technology Committee Computational Biology Special Project (YDZX20233100004032 to Cheng Luo), and (24JS2830200 to Heng Xu), the Lingang Laboratory (LG-QS-202204-07 to Hao Zhang), and the Shanghai Municipal Education Commission AI for Science Project (301-0406 to Hao Zhang).

Conflict of interest

The authors declare that they have no conflict of interest.

References

[1] Husted AS, Trauelsen M, Rudenko O, Hjorth SA, Schwartz TW. 2017. GPCR-mediated signaling of metabolites. Cell Metabolism 25: 777-796

[2] Qiu S, Cai Y, Yao H, Lin C, Xie Y, et al. 2023. Small molecule metabolites: discovery of biomarkers and therapeutic targets. Signal Transduction and Targeted Therapy 8: 132

[3] Palermo A. 2023. Metabolomics- and systems-biology-guided discovery of metabolite lead compounds and druggable targets. Drug Discovery Today 28: 103460

[4] Luzarowski M, Skirycz A. 2019. Emerging strategies for the identification of protein−metabolite interactions. Journal of Experimental Botany 70: 4605-4618

[5] Cox MA, Bassi C, Saunders ME, Nechanitzky R, Morgado-Palacin I, et al. 2020. Beyond neurotransmission: acetylcholine in immunity and inflammation. Journal of Internal Medicine 287: 120-133

[6] Kopec AM, Smith CJ, Bilbo SD. 2019. Neuro-immune mechanisms regulating social behavior: dopamine as mediator? Trends in Neurosciences 42: 337-348

[7] Ye D, Xu H, Tang Q, Xia H, Zhang C, et al. 2021. The role of 5-HT metabolism in cancer. Biochimica et Biophysica Acta (BBA) - Reviews on Cancer 1876: 188618

[8] Myburgh J. 2010. Norepinephrine: more of a neurohormone than a vasopressor. Critical Care 14: 196

[9] Piazza I, Kochanowski K, Cappelletti V, Fuhrer T, Noor E, et al. 2018. A map of protein-metabolite interactions reveals principles of chemical communication. Cell 172: 358-372.e23

[10] Li X, Gianoulis TA, Yip KY, Gerstein M, Snyder M. 2010. Extensive in vivo metabolite-protein interactions revealed by large-scale systematic analyses . Cell 143: 639-650

[11] Qin W, Yang F, Wang C. 2020. Chemoproteomic profiling of protein−metabolite interactions. Current Opinion in Chemical Biology 54: 28-36

[12] Nicholson JK, Lindon JC. 2008. Metabonomics. Nature 455: 1054-1056

[13] Xu H, Zhao H, Ding C, Jiang D, Zhao Z, et al. 2023. Celastrol suppresses colorectal cancer via covalent targeting peroxiredoxin 1 . Signal Transduction and Targeted Therapy 8: 51

[14] Cheng F, Zhou Y, Li W, Liu G, Tang Y. 2012. Prediction of chemical-protein interactions network with weighted network-based inference method. PLoS One 7: e41064

[15] Lounkine E, Keiser MJ, Whitebread S, Mikhailov D, Hamon J, et al. 2012. Large-scale prediction and testing of drug activity on side-effect targets. Nature 486: 361-367

[16] Barabási AL, Gulbahce N, Loscalzo J. 2011. Network medicine: a network-based approach to human disease. Nature Reviews Genetics 12: 56-68

[17] Chen B, Butte AJ. 2016. Leveraging big data to transform target selection and drug discovery. Journal of Clinical Pharmacology & Therapeutics 99: 285-297

[18] Zhang Y, Liu C, Liu M, Liu T, Lin H, et al. 2023. Attention is all you need: utilizing attention in AI-enabled drug discovery. Briefings in Bioinformatics 25: bbad467

[19] Wishart DS, Guo A, Oler E, Wang F, Anjum A, et al. 2022. HMDB 5.0: the human metabolome database for 2022. Nucleic Acids Research 50: D622-D631

[20] Gilson MK, Liu T, Baitaluk M, Nicola G, Hwang L, et al. 2016. BindingDB in 2015: a public database for medicinal chemistry, computational chemistry and systems pharmacology. Nucleic Acids Research 44: D1045-D1053

[21] Schriml LM, Munro JB, Schor M, Olley D, McCracken C, et al. 2022. The human disease ontology 2022 update. Nucleic Acids Research 50: D1255-D1261

[22] Gargano MA, Matentzoglu N, Coleman B, Addo-Lartey EB, Anagnostopoulos AV, et al. 2024. The human phenotype ontology in 2024: phenotypes around the world. Nucleic Acids Research 52: D1333-D1346

[23] Zdrazil B, Felix E, Hunter F, Manners EJ, Blackshaw J, et al. 2024. The ChEMBL Database in 2023: a drug discovery platform spanning multiple bioactivity data types and time periods. Nucleic Acids Research 52: D1180-D1192

[24] Yin J, Chen KM, Clark MJ, Hijazi M, Kumari P, et al. 2020. Structure of a D2 dopamine receptor−G-protein complex in a lipid membrane. Nature 584: 125-129

[25] Xu P, Huang S, Krumm BE, Zhuang Y, Mao C, et al. 2023. Structural genomics of the human dopamine receptor system. Cell Research 33: 604-616

[26] Brzozowski AM, Pike ACW, Dauter Z, Hubbard RE, Bonn T, et al. 1997. Molecular basis of agonism and antagonism in the oestrogen receptor. Nature 389: 753-758

[27] Toy W, Shen Y, Won H, Green B, Sakr RA, et al. 2013. ESR1 ligand-binding domain mutations in hormone-resistant breast cancer. Nature Genetics 45: 1439-1445

[28] Fortunati N, Catalano MG, Boccuzzi G, Frairia R. 2010. Sex Hormone-Binding Globulin (SHBG), estradiol and breast cancer. Molecular and Cellular Endocrinology 316: 86-92

[29] Maggiolini M, Vivacqua A, Fasanella G, Recchia AG, Sisci D, et al. 2004. The G protein-coupled receptor GPR30 mediates c-fos up-regulation by 17β-estradiol and phytoestrogens in breast cancer cells. Journal of Biological Chemistry 279: 27008-27016

[30] Wang N, He X, Zhao J, Jiang H, Cheng X, et al. 2022. Structural basis of leukotriene B4 receptor 1 activation. Nature Communications 13: 1156

[31] Grishkovskaya I, Avvakumov GV, Sklenar G, Dales D, Hammond GL, et al. 2000. Crystal structure of human sex hormone-binding globulin: steroid transport by a laminin G-like domain. The EMBO Journal 19: 504-512

[32] Xing C, Zhang J, Zhao H, He B. 2022. Effect of sex hormone-binding globulin on polycystic ovary syndrome: mechanisms, manifestations, genetics, and treatment. International Journal of Women's Health 14: 91-105

[33] Bhasin S, Brito JP, Cunningham GR, Hayes FJ, Hodis HN, et al. 2018. Testosterone therapy in men with hypogonadism: an endocrine society clinical practice guideline. The Journal of Clinical Endocrinology & Metabolism 103: 1715-1744

[34] Terawaki K, Yokomizo T, Nagase T, Toda A, Taniguchi M, et al. 2005. Absence of leukotriene B4 receptor 1 confers resistance to airway hyperresponsiveness and Th2-type immune responses. The Journal of Immunology 175: 4217-4225

[35] Sumida H, Yanagida K, Kita Y, Abe J, Matsushima K, et al. 2014. Interplay between CXCR2 and BLT1 facilitates neutrophil infiltration and resultant keratinocyte activation in a murine model of imiquimod-induced psoriasis. The Journal of Immunology 192: 4361-4369

[36] Zhou J, Lai W, Yang W, Pan J, Shen H, et al. 2018. BLT1 in dendritic cells promotes Th1/Th17 differentiation and its deficiency ameliorates TNBS-induced colitis. Cellular & Molecular Immunology 15: 1047-1056

[37] Mauvais-Jarvis F, Bhasin S. 2026. Metabolic messengers: testosterone. Nature Metabolism 8: 52-61

[38] He R, Chen Y, Cai Q. 2020. The role of the LTB4-BLT1 axis in health and disease. Pharmacological Research 158: 104857

[39] Zheng L, Cao J, Jing L, Kang D, Wang Z, et al. 2026. Convergence of computer-aided drug discovery and artificial intelligence: towards next-generation therapeutics. Pharmaceutical Science Advances 4: 100100

[40] Yang X, Zhang B, Wang S, Lu Y, Chen K, et al. 2023. OTTM: an automated classification tool for translational drug discovery from omics data. Briefings in Bioinformatics 24: bbad301

[41] Lamb J, Crawford ED, Peck D, Modell JW, Blat IC, et al. 2006. The connectivity map: using gene-expression signatures to connect small molecules, genes, and disease. Science 313: 1929-1935

[42] Swinney DC, Anthony J. 2011. How were new medicines discovered? Nature Reviews Drug Discovery 10: 507-519

[43] Eder J, Sedrani R, Wiesmann C. 2014. The discovery of first-in-class drugs: origins and evolution. Nature Reviews Drug Discovery 13: 577-587