Artificial intelligence (AI) researchers and cheminformatics specialists strive to identify effective drug precursors while optimizing costs and accelerating development processes. Digital molecular representation plays a crucial role in achieving this objective by making molecules machine-readable, thereby enhancing the accuracy of molecular prediction tasks and facilitating evidence-based decision making. This study presents a comprehensive review of small molecular representations and AI-driven drug discovery downstream tasks utilizing these representations. The research methodology begins with the compilation of small molecule databases, followed by an analysis of fundamental molecular representations and the models that learn these representations from initial forms, capturing patterns and salient features across extensive chemical spaces. The study then examines various drug discovery downstream tasks, including drug-target interaction (DTI) prediction, drug-target affinity (DTA) prediction, drug property (DP) prediction, and drug generation, all based on learned representations. The analysis concludes by highlighting challenges and opportunities associated with machine learning (ML) methods for molecular representation and improving downstream task performance. Additionally, the representation of small molecules and AI-based downstream tasks demonstrates significant potential in identifying traditional Chinese medicine (TCM) medicinal substances and facilitating TCM target discovery.
Artificial Intelligence ; Drug Discovery/methods* ; Humans ; Machine Learning ; Medicine, Chinese Traditional ; Small Molecule Libraries/chemistry*

Metabolic dysfunction-associated steatotic liver disease (MASLD), the most prevalent chronic liver condition globally, lacks adequate and effective therapeutic remedies in clinical practice. Recent studies have increasingly highlighted the close connection between the ubiquitin-proteasome system (UPS) and the progression of MASLD. This relationship is crucial for understanding the disease's underlying mechanism. As a sophisticated process, the UPS govern protein stability and function, maintaining protein homeostasis, thus influencing a multitude of elements and biological events of eukaryotic cells. It comprises four enzyme families, namely, ubiquitin-activating enzymes (E1), ubiquitin-conjugating enzymes (E2), ubiquitin-protein ligases (E3), and deubiquitinating enzymes (DUBs). This review aims to delve into the array of pathways and therapeutic targets implicated in the ubiquitination within the pathogenesis of MASLD. Therefore, this review unveils the role of ubiquitination in MASLD while spotlighting potential therapeutic targets within the context of this disease.

Acute liver injury (ALI) serves as a critical precursor and major etiological factor in the progression and ultimate manifestation of various hepatic disorders. The prevention and treatment of ALI is still a serious global challenge. Given the limited therapeutic options for ALI, exploring novel targeted therapeutic agents becomes imperative. The potential therapeutic efficacy of inhibiting RIPK2 is highlighted, as it may provide significant benefits by attenuating the MAPK pathway and NF-κB signaling. Herein, we propose a CMD-OPT model, a two-stage molecular optimization tool for the rapid discovery of RIPK2 inhibitors with optimal properties. Compound RP20, which targets the ATP binding site, demonstrated excellent kinase specificity, ideal oral pharmacokinetics, and superior therapeutic effects in a model of APAP-induced ALI, positioning RP20 as a promising preclinical candidate. This marks the first application of RIPK2 inhibitors in ALI treatment, opening a novel therapeutic pathway for clinical applications. These results highlight the efficacy of the CMD-OPT model in producing lead compounds from known active molecules, showcasing its significant potential in drug discovery.

Objective To construct an evidence-based discharge preparation service plan for patients after receiving percutaneous transhepatic biliary drainage(PTBD)so as to provide a theoretical reference basis for improving the quality of discharge nursing service.Methods A computerized retrieval of academic papers concerning the discharge preparation service plan for patients after receiving PTBD was conducted.The quality of the included literature was evaluated and the evidences were summarized.According to the clinical actual requirements,the first draft of discharge preparation service for patients after receiving PTBD was formed.Using Delphi method,two rounds of letter inquiries were conducted in 17 experts to determine the final version.Results In the first round of expert consultation,17 questionnaires were distributed and 15 questionnaires were recovered;and in the second round of expert consultation,15 questionnaires were distributed and 15 questionnaires were recovered.In the first round of expert correspondence,11 experts made suggestions for modification,and in the second round of expert correspondence,4 experts made suggestions,indicating that the experts were more motivated to participate in the research.The coefficient of expert consultation judgment(Ca)was 0.90,the degree of familiarity(Cs)was 0.91,the coefficient of authority(Cr)was 0.91,and the Kendall's w for round 1 and round 2 were 0.363 and 0.368 respectively.The final discharge preparation service scheme consisted of 13 items at six different time points from patient admission to after discharge.Conclusion The established discharge preparation service plan for patients after receiving PTBD is scientific and reliable,which can provide theoretical basis for patients'discharge service.

Objective Based on the best evidences to establish the practice plan of discharge preparation service for patients after receiving percutaneous transhepatic biliary drainage(PTBD),and to assess its clinical application value.Methods According to the PIPOST principle the clinical questions were proposed,the best evidences of discharge preparation service for patients after receiving PTBD were retrieved and summarized.The review indicators and review methods were formulated.The baseline review was carried out,the facilitators and barrier factors were analyzed,the change strategies were developed,the clinical transforms were implemented,and the patient outcomes were evaluated.Results After application of the evidences,the implementation of the review indicators of discharge preparation services after PTBD was improved.After discharge,the incidence of catheter complications(including catheter falling-off and puncture site skin infection)was decreased,and the difference was statistically significant(P＜0.05).There were no statistically significant differences in the incidences of tube obstruction and fluid extravasation.Conclusion The evidence-based practice of discharge preparation service for patients after receiving PTBD is helpful for improving the self-care ability of patients after discharge,reducing the incidence of tubular complications and improving the clinical outcome of patients.

Objective To explore the risk factors for bleeding at the puncture site after femoral artery puncture intervention.Methods A computerized retrieval of observation studies,including cross-sectional studies,case-control studies,and cohort studies,about the risk factors for bleeding at the puncture site after femoral artery puncture intervention from the databases of China National Knowledge Infrastructure(CNKI),Wanfang,VIP,China Biomedical Literature Service(CBM),PubMed,Medline,The Cochrane Library,EMbase and Web of Science was conducted.The retrieval time period was from the establishment of the database to December 31,2022.Newcastle-Ottawa scale(NOS)was used to evaluate the quality of the included studies,and RevMan5.3 software was used to make meta-analysis of the literature data.Results A total of 8 articles with a total sample size of 35 250 patients were included in this analysis.There were 1 410 patients in the postoperative bleeding group and 33 840 patients in the non-bleeding group.The results of the meta-analysis showed that the aged(OR=2.71,95％ CI=2.17-3.38),female(OR=4.26,95％ CI=1.08-16.89),hypertension(OR=2.48,95％CI=1.69-3.63),obesity(OR=2.33,95％ CI=1.59-3.42),use of thrombolytic agents,anticoagulants or platelet antagonists(OR=2.95,95％ CI=2.24-3.89),manual compression(OR=6.78,95％ CI=1.34-34.43)were the risk factors for the bleeding at the puncture site after femoral artery puncture intervention.Conclusion The aged,female,hypertension,obesity,use of thrombolytic agents,anticoagulants or platelet antagonists,and manual compression are the risk factors for the bleeding at the puncture site after femoral artery puncture intervention.(J Intervent Radiol,2024,33:314-320)

Purpose To investigate the value of a radiomics model based on dynamic contrast-enhanced magnetic resonance imaging(DCE-MRI)in the differential diagnosis of granulomatous mastitis and breast cancer.Materials and Methods The MRI data of 82 patients with granulomatous mastitis or breast cancer confirmed by pathology in Xiyuan Hospital of China Academy of Chinese Medical Sciences from February 2019 to January 2022 were retrospectively collected.Based on the first phase of DCE-MRI,the regions of interest(ROI)were delineated layer by layer by semi-automatic segmentation method and manual segmentation method,respectively.99 ROI were randomly assigned to 69 in training groups and 30 in test groups.The consistency difference between the two methods was compared.The original data extracted by the semi-automatic segmentation method were screened by correlation analysis and multi-factor Logistic regression.Six kinds of classifiers(Logistic regression,support vector machine,naive Bayes,decision tree,random forest,K nearest neighbor)were used to construct prediction models,and the differences in diagnostic efficiency,accuracy,sensitivity and specificity of each model were evaluated.Results A total of 99 lesions(n=37 cases with granulomatous mastitis and n=62 cases with breast cancer)were segmented from 82 patients.The radiomics data extracted by the two ROI segmentation methods had poor consistency between groups[Intraclass correlation coefficient=0.68(0.51,0.78)].Among the six prediction models constructed from the data extracted by the semi-automatic segmentation method,the diagnostic performance of the Logistic regression model and the support vector machine model was significantly better than those of other models,and the Logistic regression model had the best diagnostic performance and stability(training group:area under the curve 0.928,accuracy rate 0.855,sensitivity 0.837,specificity 0.885;test group:area under the curve 0.933,accuracy 0.833,sensitivity 0.895,specificity 0.727,respectively).Conclusion Radiomics based on DCE-MRI can provide high value for the differential diagnosis of granulomatous mastitis and breast cancer.The semi-automatic segmentation method is more recommended for the segmentation method of ROI.The prediction model constructed by Logistic regression and support vector machine shows better diagnostic efficiency and stability.

Objective:To quantitatively analyze the plaque components of carotid artery through energy spectrum computed tomography angiography(CTA),and to measure the blood flow perfusion in the blood-supply area of carotid artery through CT perfusion(CTP),so as to explore the relationship among plaque component,the degree of luminal stenosis and cerebral blood flow perfusion.Methods:A total of 68 patients with unilateral plaques of carotid artery and severe vascular stenosis who were screened and diagnosed by ultrasound and CTA in Xiyuan Hospital from December 2017 to July 2019 were selected,and all patients underwent CTA examination and CTP examination.North American symptomatic carotid endarterectomy test(NASCET)method was used to measure the degree of carotid stenosis.The GE AW 4.7 post-process workstation was used to conduct analyses of energy spectrum and cerebral perfusion for the plaque component.And then,the slope of energy spectrum curve and the effective atomic number were obtained.At the same time,the cerebral blood volume(CBV),cerebral blood flow(CBF),time to peak(TTP)and mean transit time(MTT)of contrast agent in blood-supplying area of anterior cerebral artery(ACA)and middle cerebral artery(MCA)at the side of lesion were measured.Results:A total of 68 measured plaques of 68 patients met the condition,including 44 vulnerable plaques(including lipid plaques and mixed plaques)and 24 stable plaques(fibrous plaques).The average slopes of the energy spectrum curves of vulnerable plaque and stable plaque were respectively 0.45±0.45 and 1.15±0.39,and the differences were significant(t=2.413,P<0.05).The averagely effective atomic numbers of vulnerable plaques and stable plaques were respectively 7.21±1.06 and 8.01±0.63,and the difference were significant(t=2.548,P<0.05).The average TTP values of the ACA at the side of lesion of vulnerable plaques and stable plaques were respectively(12.20±1.61)S and(13.59±2.79)S,and the difference was significant(t=-2.607,P<0.05).The mean MTT values of the ACA at the side of lesion of vulnerable plaques and stable plaques were respectively(5.07±1.66)S and(6.09±2.19)S,and the difference was significant(t=-2.177,P<0.05).The degree of vascular stenosis at the side of lesion was positively correlated with TTP and MTT in blood-supplying area of middle cerebral artery(MCA)at the side of lesion(r=0.537,0.465,P<0.05),and that was negatively correlated with CBF values in blood-supplying areas of ACA and MCA at the side of lesion(r=-0.281,-0.569,P<0.05),respectively.The slope of the energy spectrum curve of carotid plaque was positively correlated with the TTP values in blood-supplying areas of ACA and MCA at the side of lesion(r=0.242,0.246,P<0.05),respectively.Conclusion:CT spectral imaging can quantitatively analyze the displayed components of carotid atherosclerotic plaque,and the degree of vascular stenosis can affect the blood flow perfusion of cerebral tissue,and the delays of TTP and MTT are more easily caused by vulnerable plaque,and the TTP of them is more sensitivity.

Objective To explore the effect of multi-sensory artificial intelligence feedback gait training on the recovery of walking function in stroke patients based on enriched environment theory. Methods From July,2021 to June,2023,a total of 80 stroke patients in Huashan Hospital Affiliated to Fudan University were randomly divided into control group(n=40)and experimental group(n=40).Both groups received rou-tine rehabilitation in the lying and seated positions,for 40 minutes.The control group received ground walking training,for 20 minutes,while the experimental group received multi-sensory feedback gait training in enriched environment,for 20 minutes.Before and after four weeks intervention,the digital motion monitoring treadmill was used to mearsure step speed,step length,hip and knee swing angle and weight symmetry.They were as-sessed with Berg Balance scale(BBS),Fugl-Meyer Assessment-Lower Extremities(FMA-LE)and Barthel Index(BI). Results After intervention,the hip swing angle,step length of both sides and step speed significantly improved in both groups(|t|＞3.162,P＜0.05),and they were better in the experimental group than in the control group(|t|＞2.568,P＜0.05);the average knee joint swing angle and bilateral weight-bearing symmetry significantly im-proved in the experimental group(|t|＞3.249,P＜0.01);the scores of BBS,FMA-LE and BI improved in both groups(|t|＞3.569,P＜0.01),and they were better in the experimental group than in the control group(|t|＞2.922,P＜0.05). Conclusion Multi-sensory feedback gait training based on enriched environment theory could effectively improve the walking and balance of stroke patients,and increase the ability of independence.

Objective:The incidence of infections in patients with malignant hematologic diseases is extremely high and significantly affects their prognosis.Identifying early and precise biomarkers for infection is crucial for guiding the treatment of infections in these patients.Previous studies have shown that procalcitonin(PCT)can serve as an early diagnostic marker for bloodstream infections in patients with malignant hematologic diseases.This study aims to compare serum PCT levels in these patients with different pathogens,disease types,infection sites,and severity levels. Methods:Clinical data and laboratory results of infected patients with malignant hematologic diseases treated at the Department of Hematology,the Third Xiangya Hospital of Central South University from January 2018 to August 2023 were collected.General patient information was retrospectively analyzed.Serum PCT levels were compared among patients with different pathogens,types of malignant hematologic diseases,infection sites,and infection severity;Receiver operator characteristic(ROC)curves were used to determine the cut-off values and diagnostic value of serum PCT levels in diagnosing bloodstream infections versus local infections and severe infections versus non-severe infections.Mortality rates after 4-7 days of anti-infective treatment were compared among groups with rising,falling,and unchanged PCT levels. Results:A total of 526 patients with malignant hematologic diseases were included.The main pathogens were Gram-negative bacteria(272 cases,51.7%),followed by Gram-positive bacteria(120 cases,22.8%),fungi(65 cases,12.4%),viruses(23 cases,4.4%),and mixed pathogens(46 cases,8.7%).The main types of malignant hematologic diseases were acute myeloid leukemia(216 cases,41.1%),acute lymphoblastic leukemia(107 cases,20.3%),and lymphoma(93 cases,17.7%).Granulocyte deficiency was present in 68.3%(359 cases)of the patients during infection,with severe infection in 24.1%(127 cases).Significant differences in serum PCT levels were found among patients with different types of pathogens(P<0.001),with the highest levels in Gram-negative bacterial infections.Significant differences in serum PCT levels were also found among patients with different types of malignant hematologic diseases(P<0.05),with the highest levels in lymphoma patients.Serum PCT levels were significantly higher in systemic infections and severe infections compared to local infections and non-severe infections(both P<0.001).ROC curve analysis showed that the cut-off values for diagnosing bloodstream infections and severe infections were 0.22 and 0.28 ng/mL,with areas under the curve of 0.670 and 0.673,respectively.After 4-7 days of anti-infective treatment,the mortality rates of the PCT declining,PCT unchanged,and PCT rising groups were 11.9%,21.2%,and 35.7%,respectively,and pairwise comparisons were statistically significant(all P<0.05). Conclusion:PCT can be used as an auxiliary indicator for early identification of different pathogens,infection sites,and severity levels in patients with malignant hematologic diseases combined with infections.Dynamic monitoring of PCT levels after empirical antibiotic treatment provides important guidance for assessing patient's prognosis.