Identification strategy of cold and hot properties of Chinese herbal medicines based on artificial intelligence and biological experiments

To the Editor: Chinese herbal medicines (CHMs) are critical components of traditional Chinese medicine (TCM),[1] with their cold and hot properties being fundamental for guiding clinical applications and ensuring therapeutic efficacy.[2] However, the traditional method of identifying the cold and hot properties of CHMs relies heavily on the subjective experience of TCM practitioners and clinical practice, often leading to uncertainty and inconsistency. As such, there is a pressing need for more precise, rapid, and objective strategies to accurately identify the cold and hot properties of CHMs and their active ingredients.[3] Cold and hot properties of Chinese herbal medicines refer to the inherent therapeutic properties of medicinal substances, reflecting their tendency to influence the balance of yin and yang, as well as the thermal states within the human body. It is one of the primary theoretical foundations for elucidating the mechanisms of CHMs action. After a preliminary study, it was found that modern research on the cold and hot properties of CHMs is often hampered by significant subjectivity, lack of standardization, and incomplete data, limiting the scientific robustness and practical applicability of the findings. To address these challenges, we proposed an innovative identification strategy that integrates artificial intelligence (AI) models with biological experiments.[4] This combined approach aims to enhance accuracy and objectivity of traditional identification by providing a data-driven systematic framework for evaluating the cold and hot properties of CHMs. Furthermore, it offers insights into how these properties influence energy metabolism, thereby supporting the modernization of CHM clinical applications and foundational research. In this study, we developed and validated an AI-based model to identify the cold and hot properties of CHMs, leveraging biological mechanisms associated with energy metabolism. The model was constructed by using molecular fingerprinting techniques and validated through in vitro cell assays and gene expression analyses, providing both computational and experimental evidence to support the accuracy of the model. Initially, given the lack of existing AI algorithms integrated with biological experiments to characterize the hot and cold properties of CHMs from a modern scientific perspective using a large-scale dataset of ingredients, we selected all CHMs labeled as either cold or hot from the 2020 edition of the Chinese Pharmacopoeia. This selection resulted in a dataset of 266 CHMs, including 148 with cold properties and 118 with hot properties. And we screened out the molecules of the ingredients that contained in cold- and hot-propertied CHMs from the Traditional Chinese Medicine Systems Pharmacology Database and Analysis Platform (TCMSP) (https://old.tcmsp-e.com), totaling 5550 ingredients of CHMs. Then, the 167-dimensional vectors in the Maccskeys molecular fingerprinting algorithm were used to characterize each ingredient of the CHMs. The Maccskeys molecular fingerprint, developed by metadata lock (MDL), incorporates 166 features with a total length of 167 bits. Here, the 0th bit serves as a placeholder, while bits 1–166 correspond to molecular substructure features. The ingredient vector features from each CHM were fused and normalized, resulting in 167-dimensional vectorized features for each cold- or hot-propertied CHM. A random segmentation method was applied to divide the cold- and hot-propertied CHMs into training, validation, and test sets. The training dataset was utilized to build a model for identifying the cold and hot properties of CHMs. The test dataset was subsequently applied to assess the model's performance, and finally, the validation dataset was employed to rigorously validate the model's accuracy and robustness. For the above three datasets, machine learning algorithms, including support vector machines (SVMs), extreme gradient boosting (XGBoost), deep neural networks (DNNs), and random forest (RF) plots, were used to construct identification models of the cold and hot properties of CHMs. The area under curve (AUC) value was used to evaluate the performance and effectiveness of each model. Among the constructed identification models, the optimal model in terms of various indices was selected as the core algorithm. Using the medicinal property identification results of cold- and hot-property CHMs in the test set from the optimal model as a reference, the active ingredients of typical cold- and hot-property CHMs were selected for further biological validation. Cell counting kit-8 (CCK-8) cell proliferation assays were conducted to assess the effects of the active ingredients of cold- and hot-propertied CHMs on cell growth. Additionally, polymerase chain reaction (PCR) gene detection assays were performed to explore the regulatory effects of these ingredients on key energy metabolism targets, and monoamine oxidase A (MAOA), hydroxyacyl-coenzyme A (CoA) dehydrogenase trifunctional multienzyme complex subunit beta (HADHB), enoyl-CoA hydratase and 3-hydroxyacyl CoA (EHHADH), uncoupling protein1 (Ucp1), adenosine 5′-monophosphate (AMP)-activated protein kinase (AMPK), and cytochrome c oxidase subunit 5A (Cox-5a) were the main targets of our research. These experimental methods further verified the accuracy of the AI-based medicinal property identification model. Integrating AI technology with energy metabolism biological experiments to develop a strategy for identifying the hot and cold properties of CHMs represents a scientifically significant and highly applicable area of study. Biological techniques, including CCK-8, PCR, and Western blot, were used to qualitatively interpret the cold and hot properties of CHMs, confirming the accuracy of the AI-based identification model from the perspective of biological mechanisms. This study innovatively identified the cold and hot properties of CHMs by integrating AI algorithms with biological experiments. Initially, 266 CHMs with known cold and hot properties were selected from the 2020 edition of the Chinese Pharmacopoeia, and 5550 ingredients were extracted from the TCMSP database. Four AI algorithms were utilized to construct models for property identification, with the RF model emerging as the most effective. The CCK-8 cell proliferation assay indicated that active ingredients of cold-propertied CHMs inhibited cell proliferation, while those of hot-propertied CHMs promoted it, suggesting a link between energy metabolism and CHM properties. PCR gene assays further revealed that cold-propertied CHM ingredients suppressed key energy metabolism targets, whereas hot-propertied CHM ingredients activated them. These results confirmed the accuracy of the AI-based CHM property identification model and underscored the feasibility and scientific validity of using AI models combined with biological experiments to determine the cold and hot properties of CHMs. These experiments also highlighted the scientific rigor and practical applicability of the identification strategy. Thus, the optimal identification model for cold and hot properties of CHMs was successfully constructed using AI algorithms, and the model's predictions were validated through biological mechanism experiments, as illustrated in Figure 1.Figure 1: A flowchart depicting the utilization of AI algorithms in conjunction with biological experiments for identifying the cold and hot properties of CHMs. AI: Artificial intelligence; AMPK: Adenosine 5′-monophosphate (AMP)-activated protein kinase; CHMs: Chinese herbal medicines; Cox-5a: Cytochrome c oxidase subunit 5A; DNN: Deep neural network; EHHADH: Enoyl-coenzyme A (CoA) hydratase and 3-hydroxyacyl CoA; HADHB: Hydroxyacyl-CoA dehydrogenase trifunctional multienzyme complex subunit beta; MAOA: Monoamine oxidase A; RF: Random forest; SVM: Support vector machine; Ucp1: Uncoupling protein1; XGBoost: Extreme gradient boosting.In this work, we focused on strategies to identify the cold and hot properties of CHMs based on AI and energy metabolism experiments. Additionally, we discuss trends, problems, and future challenges in this field. In recent years, the application of AI technology in the identification of cold and hot properties of CHMs has gained momentum. When combined with biological experiments, AI can further enhance the accuracy and efficiency of identifying these properties. Future development trends primarily include two aspects: the application of deep learning algorithms and multisource data fusion. Deep learning has achieved remarkable results in the fields of natural language processing and image recognition, and it shows great potential in the field of identifying the cold and hot properties of CHMs. By training models on large amounts of CHMs data, these models are expected to identify the cold and hot properties of CHMs more accurately. Meanwhile, combining multiple data sources, such as the chemical composition of CHMs, results of energy metabolism experiments, and clinical efficacy data, will improve the accuracy of identifying the cold and hot properties of CHMs. The fusion of data from multiple sources will help uncover the correlations between the properties of CHMs, thereby advancing our understanding of their cold and hot attributes. Although progress has been made in identifying the cold and hot properties of CHMs based on AI and energy metabolism experiments, there are still some problems with data quality and model generalization capability. High-quality data form the basis for training accurate models. However, the quality of current CHM data is inconsistent, with many datasets containing incomplete or erroneous information. Therefore, strict data standards and quality control mechanisms must be established to improve the accuracy of identifying the cold and hot properties of CHMs. Furthermore, while current AI models perform well on training datasets, their generalizability to new CHMs and unknown contexts needs further optimization. This necessitates ongoing improvements to enhance the models' robustness in real-world applications. In summary, the combination of AI algorithms and biological mechanism experiments provides an effective strategy for identifying the cold and hot properties of CHMs. With the continuous development of AI and experimental research on energy metabolism biology, TCM practitioners will be able to reference not only textual descriptions of the hot and cold properties of CHMs but also scientifically grounded, dynamic explanations. Further refinement of this approach and deeper exploration of the relationship between medicinal properties and energy metabolism will offer valuable scientific insights for the future research and clinical application of CHMs. Funding This work has been supported by the Scientific and Technological Innovation Project of China Academy of Chinese Medical Sciences (CACMS) (No. CI2023C065YLL), Scientific and technological innovation project of CACMS (No. CI2021B003), CACMS Innovation Fund (No. CI2021A01509), and National Natural Science Foundation of China (No. 61872297). Conflicts of interest None.