Yiyu Cheng

Background: Artificial intelligence (AI), particularly large language models (LLMs), have provided powerful tools for systematically modeling the complexity of traditional Chinese medicine (TCM). To overcome the limitations of subjective formula design and unweighted target prioritization, we developed TCMNet, an AI-powered strategy that integrates LLM-assisted disease knowledge mining, protein–protein interaction (PPI) networks, and deep learning-based binding prediction to support herbal formula evaluation and active compounds identification. Methods: TCMNet integrates AI-guided literature analysis with weighted PPI network evaluation. Parkinson’s disease (PD) was chosen as the representative case study. Disease-associated protein targets were semantically weighted using TCMChat, a TCM-specific LLM that extracts relevant targets from Chinese and English literature. Furthermore, herb-specific information data, such as composition ratios, compound abundance, and compound-protein interaction probabilities, was used to generate weighted herb-related proteins. These weights were incorporated into a PPI network to assign biological weight to each node. Four classical TCM formulas (Tianma Gouteng Decoction, Liuwei Dihuang, Qianzheng San, Dabuyin Wan), the clinically optimized Pingchan Granule (PCG), and their integrative combinations with Western medicine (Levodopa) were systematically evaluated alongside the single-herb Ginkgo biloba . Therapeutic relevance was assessed using network-based metrics such as target coverage, Jaccard similarity, and weighted proximity, with statistical significance measured by Z-scores. To validate key active compounds, we employed Boltz-2, a state-of-the-art deep learning method, to predict the binding probabilities between herbal compounds and prioritized PD-associated proteins. Results: Weighted proximity metrics markedly outperformed unweighted measures across all four evaluated TCM formulas, demonstrating the substantial benefit of integrating node weights. Among the evaluated formulas, Tianma Gouteng Decoction demonstrated superior performance in target coverage and network proximity, aligning well with existing literature. Computational validation on Pingchan Granule (PCG) confirmed that TCMNet successfully captures the therapeutic retention of formula optimization. Furthermore, integrative strategies combining TCM with Levodopa exhibited significantly enhanced network proximity compared to monotherapy, supporting the rationale for combined treatment. Moreover, the case study identified flavonoids and isoflavonoids from Ginkgo biloba as the primary bioactive constituents contributing to anti-PD activity. Boltz-2 deep learning predictions further confirmed that flavonoid compounds exhibited significantly higher binding probabilities and affinities toward key PD-associated proteins compared to non-flavonoids, thus validating the results of TCMNet. Conclusions: By explicitly incorporating protein weights through combining LLM-guided target identification with node-weighted evaluation, TCMNet offers a new AI-driven strategy for optimizing TCM. This approach enables herbal formula evaluation and optimization as well as the identification of bioactive constituents, advancing the modernization of herbal medicine research.

Autoimmune diseases (AIDs) are characterized by a breakdown in immune tolerance, wherein the patient’s immune system fails to recognize self‐tissues and subsequently attacks the body’s organs and tissues. Although there are many therapeutic medicines targeting the pathogenic mechanisms, there is currently a lack of curative or long‐term symptom control options. Chimeric antigen receptor (CAR) cells are engineered cells that express multifunctional synthetic receptors. These CAR cells specifically target killer pathogenic immune cells and offer a novel approach to treating AIDs by modulating the immune microenvironment. Recent studies have demonstrated satisfactory outcomes with this method in managing autoimmune conditions. In our review paper, we provide a comprehensive summary of previous cases utilizing CAR cell therapy for AIDs. We hope that our review will provide clinicians with more options for the treatment of AIDs.

Projection-based three-dimensional bioprinting offers an approach for manufacturing biomimetic tissues with complex spatial structures and bioactivity, presenting potential for creating implantable organs or organoids to test drug response. Nevertheless, the extended printing times required for organ-scale manufacturing represents a challenge. Here we provide step-by-step instructions to manufacture organ-scale structures using bioinks while preserving high bioactivity. This approach incorporates Ficoll 400 to mitigate the heterogeneity of bioink with respect to refractive index and density, while 4-(2-aminoethyl)benzenesulfonyl fluoride and oil-sealing ensure the stability of the bioink components, thereby allowing extended printing times. This procedure also enables high-cell-viability printing via the calibration of the pH value of the bioink. This Protocol is appropriate for users with basic laboratory skills and fundamental knowledge in biotechnology to manufacture organ-scale structures for utilization in a wide variety of experimental designs. The approach is generalizable, as demonstrated by the successful printing of corpora cavernosa structures with a cell density of 10 million per milliliter, measuring 10 mm × 10 mm × 10 mm. After 7 d of culture, the cell viability was measured at 82.5%, highlighting the potential applicability in tissue engineering. All bioink preparation and printing steps are expected to take 5 h, while the development of printed structures requires 7 d of continuous culture.

Large-scale drug discovery and repurposing is challenging. Identifying the mechanism of action (MOA) is crucial, yet current approaches are costly and low-throughput. Here we present an approach for MOA identification by profiling changes in mitochondrial phenotypes. By temporally imaging mitochondrial morphology and membrane potential, we established a pipeline for monitoring time-resolved mitochondrial images, resulting in a dataset comprising 570,096 single-cell images of cells exposed to 1,068 United States Food and Drug Administration-approved drugs. A deep learning model named MitoReID, using a re-identification (ReID) framework and an Inflated 3D ResNet backbone, was developed. It achieved 76.32% Rank-1 and 65.92% mean average precision on the testing set and successfully identified the MOAs for six untrained drugs on the basis of mitochondrial phenotype. Furthermore, MitoReID identified cyclooxygenase-2 inhibition as the MOA of the natural compound epicatechin in tea, which was successfully validated in vitro. Our approach thus provides an automated and cost-effective alternative for target identification that could accelerate large-scale drug discovery and repurposing.

A pressing challenge in spatially resolved transcriptomics (SRT) is to benchmark the computational methods. A widely-used approach involves utilizing simulated data. However, biases exist in terms of the currently available simulated SRT data, which seriously affects the accuracy of method evaluation and validation. Herein, we present scCube (https://github.com/ZJUFanLab/scCube), a Python package for independent, reproducible, and technology-diverse simulation of SRT data. scCube not only enables the preservation of spatial expression patterns of genes in reference-based simulations, but also generates simulated data with different spatial variability (covering the spatial pattern type, the resolution, the spot arrangement, the targeted gene type, and the tissue slice dimension, etc.) in reference-free simulations. We comprehensively benchmark scCube with existing single-cell or SRT simulators, and demonstrate the utility of scCube in benchmarking spot deconvolution, gene imputation, and resolution enhancement methods in detail through three applications.