中草药

基于BiGRU-复合时序注意力网络的补阳还五汤多成分多靶点动态时空响应机制研究

PDF下载

中草药 | 数据挖掘与循证医学 2026,57(18): 7347-7358

收起

中草药 |数据挖掘与循证医学 2026 , 57 (18) : 7347 -7358

基于BiGRU-复合时序注意力网络的补阳还五汤多成分多靶点动态时空响应机制研究

全屏

刘灿^1,2,3, 贺鹏¹, 谯茹^1,2,3, 何文君^1,2, 贺福元^1,2,3

作者信息

通讯作者:

贺福元

作者简介:

刘灿: 刘灿，博士研究生，从事超分子机制与数理特征化研究。E-mail:liucan@stu.hnucm.edu.cn

Dynamic spatiotemporal response mechanisms of multi-component and multi-target effects of Buyang Huanwu Decoction based on BiGRU-integrated temporal attention network

LIU Can^1,2,3, HE Peng¹, QIAO Ru^1,2,3, HE Wenjun^1,2, HE Fuyuan^1,2,3

Affiliations

doi: 10.7501/j.issn.0253-2670.2026.18.022

文章导航

摘要

收起

目的建立基于双向门控循环单元（bidirectional gated recurrent unit，BiGRU）与时序注意力机制的可解释深度学习模型，解析补阳还五汤干预脑缺血过程中多成分-多靶点的动态时空响应机制。方法构建大鼠大脑中动脉闭塞（middle cerebral artery occlusion，MCAO）模型，于给药后多个时间点采集血浆和脑组织样本，分别采用UPLC-MS/MS测定14种补阳还五汤活性成分血药浓度，ELISA检测15个脑组织关键靶点表达水平，构建多成分-多靶点时序数据集。基于滑动窗口策略建立BiGRU-时序注意力网络，并结合留一批次交叉验证（leave-one-batch-out cross-validation，LOBO-CV）及Bootstrap重抽样评价模型性能与稳定性，同时解析模型提示的高贡献时间点、潜在相关成分及响应靶点特征。结果模型在LOBO-CV中表现出良好的预测性能，各测试集R²均大于0.80，均方误差（mean squared error，MSE）均小于0.20。时间注意力分析识别给药后10 h为模型关注度较高的时间节点（注意力峰值为0.602 6）；梯度归因分析筛选出丹皮酚、毛蕊异黄酮和芒柄花黄素为主要驱动成分，脑源性神经营养因子（brain derived neurotrophic factor，BDNF）、蛋白激酶G II型（protein kinase G type II，PKG2）和钙调蛋白依赖性蛋白激酶II（calmodulin-dependent protein kinase II，CAMK2）为关键响应靶点。Bootstrap重抽样结果显示，10 h关键时间窗的识别频率达到90%，提示模型具有较好的稳定性。结论建立了基于BiGRU与时序注意力机制的补阳还五汤多成分-多靶点动态时序解析方法，尝试为补阳还五汤抗脑缺血作用机制研究提供新的时序证据，并为中药复方动态药理机制研究提供参考。

关键词

补阳还五汤 / 缺血性卒中 / 双向门控循环单元 / 时序注意力 / 深度学习 / 丹皮酚 / 毛蕊异黄酮 / 芒柄花黄素

Abstract

收起

Objective To establish an interpretable deep learning model based on a bidirectional gated recurrent unit (BiGRU) and a temporal attention mechanism to elucidate the dynamic spatiotemporal response mechanisms of the multi-component–multi-target interactions underlying the effects of Buyang Huanwu Decoction (BYHWD) against ischemic stroke (IS). Methods A rat model of middle cerebral artery occlusion (MCAO) was established. Plasma and brain tissue samples were collected at multiple time points after administration. Plasma concentrations of 14 active components of BYHWD were quantified by UPLC–MS/MS, and the expression levels of 15 key targets in brain tissue were determined by ELISA to construct a multi-component–multi-target time series dataset. A BiGRU-based temporal attention network was established using a sliding window strategy, and leave-one-batch-out cross-validation (LOBO-CV) combined with bootstrap resampling was employed to evaluate model performance and stability. The model was further interpreted to identify highly contributive time points, potential key components, and responsive target features. Results The model demonstrated good predictive performance in LOBO-CV, with an R² value greater than 0.80 and a mean squared error (MSE) below 0.20 across all test sets. Temporal attention analysis identified 10 h post-administration as a time point receiving high model attention, with an attention peak value of 0.602 6. Gradient attribution analysis identified paeonol, calycosin, and formononetin as the major driving components, while brain derived neurotrophic factor (BDNF), protein kinase G type II (PKG2), and calmodulin-dependent protein kinase II (CAMK2) were identified as the key responsive targets. Bootstrap resampling showed that the 10 h critical time window was identified with a frequency of 90%, indicating good model stability. Conclusion This study established a BiGRU and temporal attention-based approach for dynamic time series analysis of the multi-component–multi-target actions of BYHWD. The proposed method provides new temporal evidence for elucidating the mechanisms underlying the anti-IS effects of BYHWD and offers a reference for investigating the dynamic pharmacological mechanisms of traditional Chinese medicine formulas.

Key words

Buyang Huanwu Decoction / ischemic stroke / bidirectional gated recurrent unit / temporal attention / deep learning / paeonol / calycosin / formononetin

引用本文

刘灿, 贺鹏, 谯茹, 何文君, 贺福元. 基于BiGRU-复合时序注意力网络的补阳还五汤多成分多靶点动态时空响应机制研究. 中草药, 2026 , 57 (18) : 7347 -7358 . DOI: 10.7501/j.issn.0253-2670.2026.18.022

LIU Can, HE Peng, QIAO Ru, HE Wenjun, HE Fuyuan. Dynamic spatiotemporal response mechanisms of multi-component and multi-target effects of Buyang Huanwu Decoction based on BiGRU-integrated temporal attention network[J]. Chinese Traditional and Herbal Drugs, 2026 , 57 (18) : 7347 -7358 . DOI: 10.7501/j.issn.0253-2670.2026.18.022

基金

收起

国家自然科学基金项目 (82274215); 湖南省自然科学基金项目 (2022JJ30453); 湖南省研究生科研创新项目 (CX20251190); 湖南中医药大学2022年度校级“揭榜挂帅”专项 (2022-XJJB001)

参考文献

收起

文献

收起

参考文献引证文献

排序方式：

钟靓, 吴敏, 林姮, 等. 全球与中国缺血性脑卒中的疾病负担变化趋势及DALYs预测(1990—2021) [J]. 中华神经外科疾病研究杂志, 2026, 20(4): 40-46.
亓鹏浩. 补阳还五汤治疗脑中风的理论溯源及研究进展[J]. 辽宁中医杂志, 2025, 52(7): 208-212.
Qin C, Yang S, Chu Y H, et al. Signaling pathways involved in ischemic stroke: Molecular mechanisms and therapeutic interventions [J]. Signal Transduct Target Ther, 2022, 7(1): 215.
Zhang X N, Li H Y, Zhao Y W, et al. Neuronal injury after ischemic stroke: Mechanisms of crosstalk involving necroptosis [J]. J Mol Neurosci, 2025, 75(1): 15.
Iordache M P, Buliman A, Costea-Firan C, et al. Immunological and inflammatory biomarkers in the prognosis, prevention, and treatment of ischemic stroke: A review of a decade of advancement [J]. Int J Mol Sci, 2025, 26(16): 7928.
Zhang J W, Niu J J, Zhao J W, et al. Deciphering the disease ecosystem of ischemic stroke via multi-omics and prospects for therapeutic strategies [J]. Front Neurosci, 2026, 20: 1740084.
吕智燚, 王飞, 何钦, 等. 补阳还五汤合黄连解毒汤加减治疗糖尿病脑病的磁共振波谱与代谢指标评估研究[J]. 中华中医药学刊, 2025, 43(10): 118-123.
Rao Y J, Li J J, Qiao R F, et al. Tetramethylpyrazine and Astragaloside IV have synergistic effects against spinal cord injury-induced neuropathic pain via the OIP5-AS1/ miR-34a/Sirt1/NF-κB axis [J]. Int Immunopharmacol, 2023, 115: 109546.
Liang Y, Deng S Q, Li Y, et al. Synergistic neuroprotection of artesunate and tetramethylpyrazine in ischemic stroke, mechanisms of blood-brain barrier preservation [J]. Int J Mol Sci, 2025, 26(16): 7979.
王春芳, 杨曜铭, 张新, 等. 基于网络药理学及分子对接技术研究中风协方II号治疗中风后遗症的机制[J]. 四川中医, 2025, 43(12): 138-148.
张鑫, 刘民, 黄艺珊, 等. 中药及其方剂药动学的发展历程和展望[J]. 中草药, 2025, 56(15): 5605-5615.
Li C, Jia W W, Yang J L, et al. Multi-compound and drug-combination pharmacokinetic research on Chinese herbal medicines [J]. Acta Pharmacol Sin, 2022, 43(12): 3080-3095.
Regazzoni F, Pagani S, Salvador M, et al. Learning the intrinsic dynamics of spatio-temporal processes through latent dynamics networks [J]. Nat Commun, 2024, 15: 1834.
Zheng W D, Hu J. Multivariate time series prediction based on temporal change information learning method [J]. IEEE Trans Neural Netw Learn Syst, 2023, 34(10): 7034-7048.
Zhang Y T, Zhou W J, Huang L J, et al. Efficient microbubble trajectory tracking in ultrasound localization microscopy using a gated recurrent unit-based multitasking temporal neural network [J]. IEEE Trans Ultrason Ferroelectr Freq Control, 2024, 71(12): 1714-1734.
Parwati F A, Wiharja K, Kurniawan I. Application of graph neural network to predict drug-drug interactions on warfarin based on molecular properties and molecular mechanism [C] // 2024 International Conference on Data Science and Its Applications (ICoDSA). Kuta, Bali, Indonesia: IEEE, 2024: 311-316.
Sun C C, Li C H, Lin X, et al. Attention-based graph neural networks: A survey [J]. Artif Intell Rev, 2023, 56(2): 2263-2310.
张朋飞, 胡蕴慧, 陈琪, 等. 网络药理学应用于医疗机构中药制剂研究与转化专家共识[J]. 中国药事, 2026, 40(5): 535-549.
Longa E Z, Weinstein PR, Carlson S, et al. Reversible middle cerebral artery occlusion without craniectomy in rats [J]. Stroke, 1989, 20(1): 84-91.
Jiang M M, Yin P L, Bai X D, et al. Proinflammatory and anti-inflammatory genes in stroke pathogenesis [J]. Curr Pharm Des, 2020, 26(34): 4220-4233.
Song K, Li Y Y, Zhang H L, et al. Oxidative stress-mediated blood-brain barrier (BBB) disruption in neurological diseases [J]. Oxid Med Cell Longev, 2020, 2020: 4356386.
Mao R, Zong N N, Hu Y J, et al. Neuronal death mechanisms and therapeutic strategy in ischemic stroke [J]. Neurosci Bull, 2022, 38(10): 1229-1247.
Duan Z D, Peng Y Q, Xu D Y, et al. Scutellarin alleviates neuronal apoptosis in ischemic stroke via activation of the PI3K/AKT signaling pathway [J]. Int J Mol Sci, 2025, 26(5): 2175.
Kuan C Y, Whitmarsh A J, Yang D D, et al. A critical role of neural-specific JNK3 for ischemic apoptosis [J]. Proc Natl Acad Sci USA, 2003, 100(25): 15184-15189.
Wang Y C, Yang Z, Wang J L, et al. Brain-body interactions in ischemic stroke: VNS reprograms microglia and FNS enhances cerebellar neuroprotection [J]. Stroke, 2025, 56(10): e267-e278.
Valable S, Montaner J, Bellail A, et al. VEGF-induced BBB permeability is associated with an MMP-9 activity increase in cerebral ischemia: Both effects decreased by Ang-1[J]. J Cereb Blood Flow Metab, 2005, 25(11): 1491-1504.
Hu Y, Zheng Y M, Wang T, et al. VEGF, a key factor for blood brain barrier injury after cerebral ischemic stroke [J]. Aging Dis, 2022, 13(3): 647-654.
Burgoyne J R, Oka S I, Ale-Agha N, et al. Hydrogen peroxide sensing and signaling by protein kinases in the cardiovascular system [J]. Antioxid Redox Signal, 2013, 18(9): 1042-1052.
Gupta A K, Gupta S, Mehan S, et al. Exploring the connection between BDNF/TrkB and AC/cAMP/PKA/ CREB signaling pathways: Potential for neuroprotection and therapeutic targets for neurological disorders [J]. Mol Neurobiol, 2025, 62(11): 14627-14659.
Tokumitsu H, Sakagami H. Molecular mechanisms underlying Ca2+/calmodulin-dependent protein kinase kinase signal transduction [J]. Int J Mol Sci, 2022, 23(19): 11025.
Peng M J, Li C, Zhang X X, et al. Research progress on the structure and physiological functions of PKG [J]. Biomed Eng Commun, 2025, 4(3): 16.

2026年第57卷第18期

PDF下载

引用本文

BibTeX

文章信息

doi: 10.7501/j.issn.0253-2670.2026.18.022

接收时间：2026-05-13
首发时间：2026-10-29

补充材料

相关文章

文章信息

作者

出版历史

收稿日期：2026-05-13

基金

作者信息

参考文献

分享链接

https://castjournals.cast.org.cn/joweb/zcy/CN/10.7501/j.issn.0253-2670.2026.18.022

分享至

全文二维码

扫描看全文

引用本文

BibTeX

本文的引用情况

2种不同金属材料的力学参数

科 Family	属数 Number of genus	种数 Number of species	占总种数比例 Percentage of total species (%)	属 Genus	种数 Number of species	占总种数比例 Percentage of total species (%)
鹅膏菌科Amanitaceae	2	11	5.26	鹅膏菌属 Amanita	10	4.78
小菇科 Mycenaceae	2	12	5.74	丝盖伞属 Inocybe	5	2.39
多孔菌科 Polyporaceae	8	14	6.70	蜡蘑属 Laccaria	5	2.39
红菇科 Russulaceae	3	23	11.00	小皮伞属 Marasmius	6	2.87
				小菇属 Mycena	11	5.26
				光柄菇属 Pluteus	5	2.39
				红菇属 Russula	17	8.13
				栓菌属 Trametes	5	2.39

关闭全屏

BibTeX
EndNote
RefWorks
TxT