OBJECTIVES: The application of photodynamic therapy in solid tumors has attracted increasing attention in recent years, and the efficiency of photosensitizers is a crucial determinant of therapeutic efficacy. This study aims to evaluate the cytotoxic effects of a novel photosensitizer, PEG-MTPABZ-PyC, in photodynamic therapy against gastric cancer cells. METHODS: Gastric cancer MKN45 cells were treated with PEG-MTPABZ-PyC. A high-content live-cell imaging system was used to assess the cellular uptake kinetics and subcellular localization of the photosensitizer. The cytotoxic effects of PEG-MTPABZ-PyC-mediated photodynamic therapy were examined using the cell counting kit-8 (CCK-8) assay and flow cytometry, while the intrinsic cytotoxicity of the photosensitizer alone was verified by the CCK-8 assay. Intracellular reactive oxygen species (ROS) generation after photodynamic therapy was detected using 2'-7'-dichlorodihydrofluorescein diacetate (DCFH-DA). RESULTS: PEG-MTPABZ-PyC alone exhibited no cytotoxicity toward MKN45 cells, indicating excellent cytocompatibility. The compound efficiently entered cells within 6 hours and localized predominantly in lysosomes. Upon light irradiation, PEG-MTPABZ-PyC-mediated photodynamic therapy induced significant cytotoxicity compared with the control group (P<0.05) and generated abundant intracellular ROS. CONCLUSIONS The novel photosensitizer PEG-MTPABZ-PyC demonstrates potent photodynamic cytotoxicity against gastric cancer cells, showing promising potential for further development in gastric cancer photodynamic therapy.
Humans ; Stomach Neoplasms/drug therapy* ; Photochemotherapy/methods* ; Photosensitizing Agents/pharmacology* ; Cell Line, Tumor ; Polyethylene Glycols/chemistry* ; Reactive Oxygen Species/metabolism* ; Mesoporphyrins/pharmacology*

Traditional Chinese medicine (TCM) represents a paradigmatic approach to personalized medicine, developed through the systematic accumulation and refinement of clinical empirical data over more than 2000 years, and now encompasses large-scale electronic medical records (EMR) and experimental molecular data. Artificial intelligence (AI) has demonstrated its utility in medicine through the development of various expert systems (e.g., MYCIN) since the 1970s. With the emergence of deep learning and large language models (LLMs), AI's potential in medicine shows considerable promise. Consequently, the integration of AI and TCM from both clinical and scientific perspectives presents a fundamental and promising research direction. This survey provides an insightful overview of TCM AI research, summarizing related research tasks from three perspectives: systems-level biological mechanism elucidation, real-world clinical evidence inference, and personalized clinical decision support. The review highlights representative AI methodologies alongside their applications in both TCM scientific inquiry and clinical practice. To critically assess the current state of the field, this work identifies major challenges and opportunities that constrain the development of robust research capabilities-particularly in the mechanistic understanding of TCM syndromes and herbal formulations, novel drug discovery, and the delivery of high-quality, patient-centered clinical care. The findings underscore that future advancements in AI-driven TCM research will rely on the development of high-quality, large-scale data repositories; the construction of comprehensive and domain-specific knowledge graphs (KGs); deeper insights into the biological mechanisms underpinning clinical efficacy; rigorous causal inference frameworks; and intelligent, personalized decision support systems.
Medicine, Chinese Traditional/methods* ; Artificial Intelligence ; Humans ; Precision Medicine ; Decision Support Systems, Clinical

Advancements in artificial intelligence (AI) and emerging technologies are rapidly expanding the exploration of chemical space, facilitating innovative drug discovery. However, the transformation of novel compounds into safe and effective drugs remains a lengthy, high-risk, and costly process. Comprehensive early-stage evaluation is essential for reducing costs and improving the success rate of drug development. Despite this need, no comprehensive tool currently supports systematic evaluation and efficient screening. Here, we present druglikeFilter, a deep learning-based framework designed to assess drug-likeness across four critical dimensions: 1) physicochemical rule evaluated by systematic determination, 2) toxicity alert investigated from multiple perspectives, 3) binding affinity measured by dual-path analysis, and 4) compound synthesizability assessed by retro-route prediction. By enabling automated, multidimensional filtering of compound libraries, druglikeFilter not only streamlines the drug development process but also plays a crucial role in advancing research efforts towards viable drug candidates, which can be freely accessed at https://idrblab.org/drugfilter/.

Drug development encompasses multiple processes, wherein protein subcellular localization is essential. It promotes target identification, treatment development, and the design of drug delivery systems. In this research, a deep learning framework called LocPro is presented for predicting protein subcellular localization. Specifically, LocPro is unique in (a) combining protein representations from the pre-trained large language model (LLM) ESM2 and the expert-driven tool PROFEAT, (b) implementing a hybrid deep neural network architecture that integrates convolutional neural network (CNN), fully connected (FC) layer, and bidirectional long short-term memory (BiLSTM) blocks, and (c) developing a multi-label framework for predicting protein subcellular localization at multiple granularity levels. Additionally, a dataset was curated and divided using a homology-based strategy for training and validation. Comparative analyses show that LocPro outperforms existing methods in sequence-based multi-label protein subcellular localization prediction. The practical utility of this framework is further demonstrated through case studies on drug target subcellular localization. All in all, LocPro serves as a valuable complement to existing protein localization prediction tools. The web server is freely accessible at https://idrblab.org/LocPro/.

Objective:To investigate the predictive value of preoperative coagulation function and inflammation response biomarkers for postoperative recurrence of non-muscle-invasive bladder cancer (NMIBC) patients.Methods:The clinical data of 390 NMIBC patients underwent surgical treatment from May 2014 to May 2021 in the Second Affiliated Hospital of Xi′an Jiaotong University were retrospectively analyzed. The baseline characteristics coagulation function, inflammation response indexes and tumor characteristics were recorded. The baseline characteristics included gender, age and smoking history; the coagulation function included prothrombin time (PT), activated partial thromboplastin time (APTT), fibrinogen (FIB) and D-dimer; the inflammation response indexes included neutrophil count, lymphocyte count, platelet count and monocyte count, and the systemic inflammatory response index (SIRI) and systemic immune-inflammation index (SII) were calculated; tumor characteristics included TNM stage, pathological grade, tumor length, tumor amount and postoperative instillation drugs. The patients were followed up until May 2022, with recurrence records and grouping. The "pROC" package was used to draw the receiver operating characteristic (ROC) curve, and calculate the optimal cutoff values of biomarkers. Multivariate Cox regression analysis was used to analyze the independent risk factors of recurrence in patients with NMIBC (variables were selected with P<0.1). The nomogram and its calibration curve were drawn by the "survival" and "rms" packages, and the area under the curve (AUC) was calculated with the "pROC" package for assessing the predictive ability of the model. The "caret" package was used for ten-fold cross-validation to evaluate the external applicability of the nomogram. Results:The ROC curve analysis result showed that the optimal cutoff values of PT, APTT, FIB, D-dimer, SIRI and SII were 11.95 s, 17.65 s, 0.233 mg/L, 565 ng/L, 0.62 and 291.5, respectively. The 390 patients with NMIBC were followed up 29 to 71 months, with a median follow-up time of 49 months. Among them, 113 patients experienced postoperative recurrence (recurrence group), and the recurrence rate was 29.0%; while 277 patients did not experience recurrence (non-recurrence group). The rate of FIB≥0.233 mg/L, D-dimmer ≥565 ng/L, SIRI≥0.62 and SII≥291.5, T 1 stage, high-grade tumor, tumor length ≥2.3 cm and multiple tumor in recurrence group were significantly higher than those in non-recurrence group: 90.3% (102/113) vs. 71.5% (198/277), 33.6% (38/113) vs. 23.5% (65/277), 74.3% (84/113) vs. 56.7% (157/277), 84.1% (95/113) vs. 60.6% (168/277), 77.9% (88/113) vs. 38.6% (107/277), 25.7% (29/113) vs. 8.3% (23/277), 49.6% (56/113) vs. 32.1% (89/277) and 41.6% (47/113) vs. 19.9% (55/277), and there were statistical differences ( P<0.01 or <0.05); there were no statistical differences in gender ratio, age, smoking history, PT, APTT and postoperative instillation drugs between the two groups ( P>0.05). Multivariate Cox regression analysis result showed that FIB≥0.233 mg/L, SII≥291.5, T 1 stage, high pathological grade, tumor length≥2.3 cm and multiple tumor were independent risk factors of postoperative recurrence in patients with NMIBC ( HR = 2.186, 1.627, 3.182, 1.675, 1.775 and 2.052; 95% CI 1.149 to 4.159, 0.913 to 2.902, 1.988 to 5.095, 1.067 to 2.630, 1.208 to 2.608 and 1.388 to 3.033; P<0.1). A nomogram model was constructed to predict postoperative 1-, 3- and 5-year non-recurrence based on FIB, SII, T stage, tumor length, pathological grade and tumor amount. The calibration curve analysis result showed that the nomogram model predicted good consistency between the postoperative 1-, 3-, 5-year non-recurrence rates and the actual incidence rate in patients with NMIBC. ROC curve analysis result showed that the AUC of the nomogram model for predicting postoperative 1-, 3- and 5-year non-recurrence in patients with NMIBC were 0.746, 0.789 and 0.835 (95% CI 0.695 to 0.832, 0.703 to 0.875 and 0.756 to 0.915). The ten-fold cross-validation result showed that the nomogram model had good external applicability for predicting postoperative 1-, 3- and 5-year non-recurrence in patients with NMIBC, with AUC of 0.754, 0.781 and 0.832 (95% CI 0.689 to 0.817, 0.724 to 0.832 and 0.778 to 0.879). Conclusions:The nomogram model based on FIB, SII, T stage, tumor length, pathological grade and tumor amount can accurately predict the postoperative 1-, 3- and 5-year recurrence risks in patients with NMIBC. The model helps clinical doctors early identify high-risk recurrent NMIBC patients, and provides reference for the development of individualized treatment plans.

Objective:To investigate the effect of blood lipids and statins use on the outcome of acute ischemic stroke (AIS) patients with cerebral microbleeds (CMBs).Methods:Consecutive AIS patients with CMBs hospitalized at the First People's Hospital of Taicang, Jiangsu Province from July 2023 to June 2024 were included retrospectively. At 3 months after onset, the modified Rankin Scale was used for outcome assessment. 0-2 was defined as good outcome and >2 was defined as poor outcome. Multivariate logistic regression analysis was used to identify independent influencing factors for poor outcome. Results:A total of 110 AIS patients with CMBs were enrolled, including 72 males (65.5%), aged 68.04±3.12 years. Thirty patients (27.3%) had poor outcome. Univariate analysis showed that age, baseline National Institutes of Health Stroke Scale (NIHSS) score, total cholesterol, triglycerides, low-density lipoprotein cholesterol, and the proportion of patients with hypertension and diabetes in poor outcome group were significantly higher than those in good outcome group, while baseline high-density lipoprotein cholesterol and the proportion of patients using statins before onset were significantly lower than those in good outcome group ( P<0.05). Multivariate logistic regression analysis showed that age (odds ratio [ OR] 1.309, 95% confidence interval [ CI] 1.007-1.702; P=0.044), the baseline NIHSS score ( OR 1.541, 95% CI 1.143-2.078; P=0.005) and high triglycerides ( OR 5.150, 95% CI 2.150-8.717; P=0.023) were the independent risk factors for poor outcome, while high high-density lipoprotein cholesterol ( OR 0.001, 95% CI 0.001-0.034; P<0.001) and statins use ( OR 0.231, 95% CI 0.046-0.558; P=0.019) were the independent protective factors for good outcome. Conclusions:Blood lipid and statins use are independent influencing factors for the outcome of AIS patients with CMBs. The use of statins before onset is associated with a lower risk of poor outcome in AIS patients with CMBs.

Drug development encompasses multiple processes,wherein protein subcellular localization is essential.It promotes target identification,treatment development,and the design of drug delivery systems.In this research,a deep learning framework called LocPro is presented for predicting protein subcellular localization.Specifically,LocPro is unique in(a)combining protein representations from the pre-trained large language model(LLM)ESM2 and the expert-driven tool PROFEAT,(b)implementing a hybrid deep neural network architecture that integrates convolutional neural network(CNN),fully connected(FC)layer,and bidirectional long short-term memory(BiLSTM)blocks,and(c)developing a multi-label framework for predicting protein subcellular localization at multiple granularity levels.Additionally,a dataset was curated and divided using a homology-based strategy for training and validation.Compar-ative analyses show that LocPro outperforms existing methods in sequence-based multi-label protein subcellular localization prediction.The practical utility of this framework is further demonstrated through case studies on drug target subcellular localization.All in all,LocPro serves as a valuable complement to existing protein localization prediction tools.The web server is freely accessible at https://idrblab.org/LocPro/.

Advancements in artificial intelligence(AI)and emerging technologies are rapidly expanding the exploration of chemical space,facilitating innovative drug discovery.However,the transformation of novel compounds into safe and effective drugs remains a lengthy,high-risk,and costly process.Comprehensive early-stage evaluation is essential for reducing costs and improving the success rate of drug development.Despite this need,no comprehensive tool currently supports systematic evaluation and efficient screening.Here,we present druglikeFilter,a deep learning-based framework designed to assess drug-likeness across four critical dimensions:1)physicochemical rule evaluated by systematic determination,2)toxicity alert investigated from multiple perspectives,3)binding affinity measured by dual-path analysis,and 4)compound synthesizability assessed by retro-route prediction.By enabling automated,multidimensional filtering of compound libraries,druglikeFilter not only streamlines the drug development process but also plays a crucial role in advancing research efforts towards viable drug candidates,which can be freely accessed at https://idrblab.org/drugfilter/.

The complement component 4 binding protein alpha(C4BPA)gene is the alpha chain of complement binding protein 4.As a plasma protein involved in the complement and coagulation systems,it can influence immune responses and lipid metabolism.In order to study the polymor-phism of C4BPA gene and its correlation with milk quality traits in Chinese Holstein cows,genom-ic DNA was extracted from blood samples of 92 Chinese Holstein cows,and the target fragment of C4BPA gene was amplified by PCR,and the association analysis was performed by using direct se-quencing to obtain the SNP loci and milk quality traits.The results showed that among the four SNPs found at the third intron of the C4BPA gene,I3-11 G＞A was highly significantly correlated with milk protein and urea nitrogen(P＜0.05),I3-291 T＞G was significantly correlated with lac-tose(P＜0.05),I3-374 C＞T was highly significantly correlated with lactose and urea nitrogen(P＜0.05),and I3-375 T＞G was highly significantly correlated with lactose(P＜0.05),milk pro-tein and urea nitrogen.The chi-square test values for each point indicated that the population was in genetic equilibrium.Individuals of haplotype combination H1 H1 had the highest lactose content,and haplotype combination H1H2 can be used as the best haplotype combination in the molecular selection work of dairy cows.

Purpose To explore the feasibility and clinical value of deep learning-based image reconstruction technology in 5.0T MRI for nasopharyngeal carcinoma.Materials and Methods A prospective study was conducted on 50 newly diagnosed nasopharyngeal carcinoma patients from August to December 2024 at Sun Yat-sen University Cancer Center.5.0T MRI was performed to scan the nasopharynx region.Routine scanning protocols included transverse T2WI,transverse T1WI,transverse contrast-enhanced T1WI and coronal fat-suppressed contrast-enhanced T1WI sequences.Based on these standard scanning protocols,DeepRecon deep learning reconstruction technology with different levels(grade 1-5)was applied,generating a total of 24 sets of images.Qualitative evaluation employed a Likert scale(5-point system)for subjective scoring on lesion detection,lesion edge clarity,artifacts and overall image quality.Quantitative evaluation was performed using the signal-to-noise ratio and contrast-to-noise ratio to objectively assess the quality of the 24 image sets.Differences in qualitative and quantitative indicators between different groups were compared,while the Kappa coefficient was used to analyze the consistency of subjective evaluations by two radiologists.Results In the qualitative assessment of 24 image sets from four MRI sequences(with and without DeepRecon reconstruction),DeepRecon images(grade 2-4)significantly outperformed traditional images in all features except for artifact reduction(Z=-12.11--6.23,all P<0.001).Images reconstructed at DeepRecon grade 3 had the highest overall score and the best image quality.Furthermore,compared with traditional images,DeepRecon images(grade 2-5)demonstrated significantly improved signal-to-noise ratio for both lesions and the lateral pterygoid muscle(t=-15.67--3.44,Z=-6.09--4.63,all P<0.01).In addition,in the transverse T2WI,transverse contrast-enhanced T1WI and coronal fat-suppressed contrast-enhanced T1WI images with DeepRecon reconstruction(grade 2-5),the contrast-to-noise ratio(lesion/lateral pterygoid muscle)also showed significant improvement compared to traditional images(t=-12.71--3.19,Z=-6.08--4.47,all P<0.001).The inter-observer agreement for the overall subjective quality score between the two radiologists was good(Kappa=0.75-0.82,all P<0.01).Conclusion DeepRecon deep learning reconstruction technology significantly increases the signal-to-noise ratio and resolution of traditional magnetic resonance images of nasopharyngeal cancer,improving image clarity and bringing more possibilities for the advancement of imaging diagnosis.