To explore the feasibility of remotely obtaining complex information on traditional Chinese medicine (TCM) pulse conditions through voice signals.

We used multi-label pulse conditions as the entry point and modeled and analyzed TCM pulse diagnosis by combining voice analysis and machine learning. Audio features were extracted from voice recordings in the TCM pulse condition dataset. The obtained features were combined with information from tongue and facial diagnoses. A multi-label pulse condition voice classification DNN model was built using 10-fold cross-validation, and the modeling methods were validated using publicly available datasets.

The analysis showed that the proposed method achieved an accuracy of 92.59% on the public dataset. The accuracies of the three single-label pulse manifestation models in the test set were 94.27%, 96.35%, and 95.39%. The absolute accuracy of the multi-label model was 92.74%.

Voice data analysis may serve as a remote adjunct to the TCM diagnostic method for pulse condition assessment.

To explore the relationship between bitter taste and bitter flavor (efficacy), their characteristics were analyzed at the target and pharmacodynamic points.

Compounds with bitter receptor activity were obtained by high-throughput screening models of the human bitter taste receptors (hTAS2Rs), which were used to analyze bitter taste. The efficacy of anti-asthma medications was used as an example to research bitter flavor (efficacy). The pharmacological effects of bitter taste and bitter flavor (efficacy) were classified according to the functional modules. The coverage for the same targets or pharmacological effects in the overall protein interaction network (PIN) of bitter taste and bitter flavor (efficacy) was analyzed to reveal their relationship. The effect of the compound polydatin with anti-asthma activation on hTAS2R14 was studied to verify the reliability of the aforementioned idea.

A total of 121 Chinese materia medica (CMM) compounds that activate hTAS2R10, hTAS2R14, and hTAS2R49 were obtained. The analysis results indicated that 108 same targets for bitter taste and bitter flavor (efficacy) were obtained, accounting for 13.9% in the PIN of bitter taste and 72.5% in the PIN of bitter flavor (efficacy). The pharmacological effects shared by bitter taste and bitter flavor (efficacy) accounted for 79% of the PIN of bitter taste and 81% of the PIN of bitter flavor (efficacy). The activating effect of the anti-asthma compound polydatin on hTAS2R14 was dose-dependent with EC₅₀ of 4.3 μM.

In this study, the relationship between bitter taste and bitter flavor (efficacy) has been demonstrated from the target and pharmacodynamic points, which are based on hTAS2Rs and anti-asthma effect. Bitter taste and bitter flavor (efficacy) exhibited a high correlation. This study provides a theoretical and scientific basis for the hypothesis that “the property theory of CMM is the clinical pharmacodynamics of CMM.”

To identify the relationship between the active compounds in turmeric (Curcuma Longa L., C. Longa) and their medicinal properties, and clarify the chemical material basis of the medicinal properties in C. Longa.

High throughput screening models of the thermo-transient receptor potentials (thermo-TRPs) and human taste type 2 receptors (hTAS2Rs) were established to evaluate the activity of the active compounds in C. Longa. The biological processes and distributions of the related targets of the active compounds were acquired by data mining. The above results were then used as basic data to comprehensively analyze the material basis of the medicinal properties of turmeric.

Curcumin inhibited TRPV1, TRPV2, TRPV3, and TRPA1, while demethoxycurcumin (DMC) inhibited TRPV3, TRPA1, and TRPM8. In terms of the 21 hTAS2Rs, the response values of hTAS2R38, hTAS2R16, and hTAS2R44 intervened by curcumin were ranked top three; the response values of hTAS2R3, hTAS2R1, and hTAS2R9 by DMC were in top three; the response values of hTAS2R47, hTAS2R39, and hTAS2R43 by bisdemethoxycurcumin (BDMC) were in top three. The biological processes related to active compounds primarily involved blood circulation, ossification, and regulation of the inflammatory response.

Curcumin, DMC, and BDMC were identified as the material basis of the property of Chinese materia medica of C. Longa. Among these, curcumin and DMC may contribute to its warming effect in the four qi. Curcumin, DMC, and BDMC may lead to its bitter flavor, and the three active compounds were consistent with the meridian entry of C. Longa. Although different compounds play diverse roles in the four qi, five flavors, and meridian entry, they all were relevant to the efficacy of C. Longa. It is helpful to further understand the important role of the property of Chinese materia medica for the clinical application of traditional Chinese medicine.

The application of Chinese materia medica (CMM) in clinical diseases is the embodiment and continuation of the property theory of CMM (PTCMM). However, due to a lack of precise quantitative description methods, it is difficult to systematically analyze the property of CMM (PCMM) and clinical effect features at the micro molecular level.

The therapeutic drugs and targets were obtained from the Drugbank database. The molecular descriptors of these drugs were calculated based on Dragon software. Drug–effect relationships that integrated the molecular descriptors and effect descriptors were plotted as grayscale images. These images were used to train the LeNet-5 model and the AlexNet model. The best-performing model was used to predict the effect features of the CMM compounds. Finally, the effect features of the PCMM combinations were calculated based on the support vector machine recursive feature elimination algorithm.

The AlexNet model showed a superior prediction performance. The results showed that its accuracy, precision, sensitivity, F-measure, and Matthews correlation coefficient on the training set were 0.940, 0.936, 0.945, 0.940, and 0.880, respectively, and those of the test set were 0.909, 0.901, 0.920, 0.910, and 0.819, respectively. A total of 399 compounds in the 42 CMMs for promoting blood circulation and removing blood stasis were predicted by this model. The key effect features of the Han-Ku-Gan combination were anti-inflammatory, anti-tumor, anti-atherosclerosis, anti-Parkinson, hypoglycemic, and anti-coagulant properties, as well as excitation of uterine smooth muscle. The key effect features of the Wen-Xin-Gan combination were anti-inflammatory, anti-atherosclerosis, anti-hypertensive, anti-coagulant, anti-tumor, and anti-cardiac insufficiency effects, as well as enhanced immunity, sedation and hypnosis, and analgesia.

This study provides a new method for the further exploration of the relationship between the PCMM and clinical effect features.

Lipid droplet (LD) deposition in adipose tissue is a critical factor leading to metabolic dysfunction. Various herbal medicines in traditional Chinese medicine (TCM) are used to treat hyperlipidemia, type 2 diabetes, obesity, and other diseases. The objective of this study was to identify potential anti-adipogenic agents from TCM herbal compounds.

One hundred and twenty compounds were evaluated in terms of their effect on adipocyte differentiation through image-based high content screening. Anti-adipogenic effects of identified hits were further confirmed at various concentrations. In addition, drug-induced liver injury assay was performed with HepG2 cells to test the hepatotoxicity of hit compounds.

Berbamine (BBM), a chemical isolated from barberry, and a derivative of BBM, berbamine dihydrochloride (BBMD), reduced LDs formation by more than 50%. Dose-dependent effects were observed and the IC₅₀ values of the two hits, BBM and BBMD, were determined as 1.88 μM and 0.95 μM, respectively. Moreover, BBM induced mild HepG2 cell injury, while its dihydrochloride—BBMD did not exhibit hepatotoxicity within 40 μM.

This study demonstrates that BBMD may be a potential therapeutic candidate for disorders associated with elevated LDs accumulation.