Jing Huang

Gastrointestinal (GI) cancers remain leading causes of cancer-related morbidity and mortality worldwide, driven by complex molecular mechanisms involving genomic, transcriptomic, proteomic, metabolomic, and radiomic alterations. The discovery of reliable biomarkers is essential for improving early detection, risk stratification, prognosis prediction, and personalized treatment strategies. Recent advances in deep learning (DL) have enabled the integration of multi-omics and imaging data, offering unprecedented opportunities to identify clinically relevant biomarkers in these malignancies. This review synthesizes representative DL architectures specifically optimized for GI cancer biomarker research, highlighting methodological advances in feature extraction, multi-modal data fusion, and predictive modeling. Beyond detailing technical architectures, we critically examine current challenges, including the reliance on public datasets, the impact of intratumoral heterogeneity, and the need for rigorous external validation. We further discuss emerging strategies such as biologically-informed deep learning and multimodal foundation models for enhancing model interpretability, addressing data heterogeneity, and supporting translational adoption within regulatory frameworks (e.g., FDA/NMPA). By bridging computational innovations with clinical needs, DL-enabled multi-omics integration holds promise for advancing predictive, preventive, and personalized medicine in gastrointestinal oncology.

Acinetobacter baumannii is a significant pathogen of hospital-acquired infections, and its multidrug resistance (MDR) and extended drug resistance (XDR) have become increasingly severe, posing a global public health challenge. This article provides a narrative review of the major resistance mechanisms of Acinetobacter baumannii, including β-lactamase production, efflux pump overexpression, target site modification, reduced membrane permeability, and biofilm formation. Additionally, it summarizes the current main drugs and their target sites for treating MDR Acinetobacter baumannii infections, with a focus on the mechanism of action, antibacterial activity, and clinical research progress of the novel fully synthetic fluorocycline antibiotic—eravacycline. Eravacycline inhibits protein synthesis by high-affinity binding to the bacterial ribosomal 30S subunit and demonstrates activity against multidrug-resistant Acinetobacter baumannii (excluding Pseudomonas aeruginosa), providing a potential novel therapeutic option for MDR/XDR Acinetobacter baumannii infections. Finally, the article outlines future research directions and treatment strategies. Due to the narrative nature of this review, no systematic methodology (e.g., PRISMA) was applied, and the available clinical evidence, particularly for CRAB infections, remains limited.

Background: Janus kinase 1 (JAK1) plays a critical role in most cytokine-mediated inflammatory, autoimmune responses and various cancers via the JAK/STAT signaling pathway. Inhibition of JAK1 is therefore an attractive therapeutic strategy for several diseases. Recently, high-performance machine learning techniques have been increasingly applied in virtual screening to develop new kinase inhibitors. Our study aimed to develop a novel layered virtual screening method based on machine learning (ML) and pharmacophore models to identify the potential JAK1 inhibitors. Methods: Firstly, we constructed a high-quality dataset comprising 3834 JAK1 inhibitors and 12,230 decoys, followed by establishing a series of classification models based on a combination of three molecular descriptors and six ML algorithms. To further screen potential compounds, we constructed several pharmacophore models based on Hiphop and receptor-ligand algorithms. We then used molecular docking to filter the recognized compounds. Finally, the binding stability and enzyme inhibition activity of the identified compounds were assessed by molecular dynamics (MD) simulations and in vitro enzyme activity tests. Results: The best performance ML model DNN-ECFP4 and two pharmacophore models Hiphop3 and 6TPF 08 were utilized to screen the ZINC database. A total of 13 potentially active compounds were screened and the MD results demonstrated that all of the above molecules could bind with JAK1 stably in dynamic conditions. Among the shortlisted compounds, the four purchasable compounds demonstrated significant kinase inhibition activity, with Z-10 being the most active (IC₅₀ = 194.9 nM). Conclusion: The current study provides an efficient and accurate integrated model. The hit compounds were promising candidates for the further development of novel JAK1 inhibitors.

Background Janus kinase 1 (JAK1) plays a critical role in most cytokine-mediated inflammatory, autoimmune responses and various cancers via the JAK/STAT signaling pathway. Inhibition of JAK1 is therefore an attractive therapeutic strategy for several diseases. Recently, high-performance machine learning techniques have been increasingly applied in virtual screening to develop new kinase inhibitors. Our study aimed to develop a novel layered virtual screening method based on machine learning (ML) and pharmacophore models to identify the potential JAK1 inhibitors. Methods Firstly, we constructed a high-quality dataset comprising 3834 JAK1 inhibitors and 12230 decoys, followed by established a series of classification models based on a combination of three molecular descriptors and six ML algorithms. To further screen potential compounds, we constructed several pharmacophore models based on Hiphop and receptor-ligand algorithms. We then used molecular docking to filter the recognized compounds. Finally, the binding stability and enzyme inhibition activity of the identified compounds were assessed by molecular dynamics (MD) simulations and in vitro enzyme activity tests. Results The best performance ML model DNN-ECFP4 and two pharmacophore models Hiphop3 and 6TPF 08 were utilized to screen the ZINC database. A total of 13 potentially active compounds were screened and the MD results demonstrated that all of the above molecules could bind with JAK1 stably in dynamic conditions. Among the shortlisted compounds, the four purchasable compounds demonstrated significant kinase inhibition activity, with Z-10 being the most active (IC₅₀ = 194.9 nM). Conclusion The current study provides an efficient and accurate integrated model. The hit compounds were promising candidates for the further development of novel JAK1 inhibitors.

To assist orthopaedic surgeons in choosing appropriate antibiotics, this study attempted to identify the common microorganisms that cause periprosthetic joint infection (PJI) and their respective drug resistance spectrums. The clinical data of 202 patients with PJI after primary hip and knee arthroplasty between January 2017 and December 2021 were retrospectively analysed. There were 84 males and 118 females, aged (63.03 ± 13.10) years (range: 24–89 years). A total of 102 and 100 patients underwent total hip and total knee arthroplasty, respectively. Based on the time of postoperative infection, the patients were divided into acute (25 cases), delayed (91 cases), and chronic (86 cases) PJI groups. The results of pathogen species, composition ratio, and drug susceptibility tests were collected. Gram‐positive bacteria were the primary causative pathogens of PJI (58.91%, 119/202), and their culture‐positive rates in patients with acute, delayed, and chronic PJI were 32.00% (8/25), 62.64% (57/91), and 62.79% (54/86), respectively. Staphylococcus epidermidis and Staphylococcus aureus were the major gram‐positive bacteria detected, followed by gram‐negative bacteria (29/202, 14.36%), and fungi (4/202, 1.98%). Gram‐positive bacteria showed higher resistance to penicillin (81.25%), oxacillin (63.33%), erythromycin (61.17%), and clindamycin (48.35%) and 100% sensitivity to linezolid, vancomycin, daptomycin, and tigecycline. In gram‐negative bacteria, the drug resistance rates to cefazolin, gentamicin, furantoin, cefuroxime, ticacillin/clavulanic acid, ceftriaxone, ciprofloxacin, and tobramycin were >50%. However, no vancomycin‐resistant bacteria were discovered in the current study. The drug resistance rate to carbapenems was low, ranging from 0% to 3.57%. Gram‐positive bacteria are the main causative pathogens of PJI, and the resistance rate of pathogens of chronic PJI is higher than those of delayed and acute PJI. Use of cefuroxime and clindamycin in patients with PJIs should proceed with caution because of the high drug resistance rate. Vancomycin can be used as a first‐line antibiotic against gram‐positive bacteria. Carbapenems can be used as the first choice against gram‐negative bacteria because of to their high sensitivity.

To investigate the application value of next‐generation sequencing (NGS) technology in the detection of pathogenic bacteria in the periprosthetic joint infection after arthroplasty. Twenty‐two cases of patients with joint infection after arthroplasty in our hospital from March 2020 to March 2021 were selected, with 11 cases of knee and 11 cases of hip, including 8 cases of male and 14 cases of female, and an average age of 63.55 ± 13.11 years old (range from 28 to 85). Microbiological culture results of synovial fluid and periprosthetic joint tissue and NGS results of periprosthetic joint tissue were collected. The detection rate of NGS and microbiological culture were calculated and statistically analysed by paired χ² test. Among the 22 patients with joint infection after arthroplasty, the positive rate of NGS was 90.91% (20/22), whereas the positive rate of bacterial culture was 50.00% (11/22). Paired chi‐square test showed a statistically significant difference in the detection rate between the two groups (P = .0029). In the detection of pathogenic microorganism, NGS detected 12 kinds of bacteria, Staphylococcus aureus in 3 patients, Staphylococcus epidermidis in 5 cases, Streptococcus 1 case, Streptococcus dysgalactiae 1 case, Xanthomonas campestris 3 cases, Escherichia coli 2 cases, Bacillus cereus 2 cases, Klebsiella pneumoniae 1 case, Finegoldia magna 1 case, Corynebacterium klopensteriella in 1 case, Brucella 1 case, and Aspergillus flavus 1 case. Bacterial culture detected 6 kinds of bacteria, included 5 cases of Staphylococcus epidermis (including 3 cases of Methicillin‐resistant coagulase‐negative Staphylococcus, (MRSCoN)), 2 cases of Staphylococcus aureus (both Methicillin‐resistant Staphylococcus aureus, (MRSA)), 1 case of Klebsiella pneumoniae, 1 case of Staphylococcus hominis (MRSCoN), 1 case of G⁺ bacillus, and 1 case of Brucella. Compared with bacterial culture, NGS technology has some advantages in the detection efficiency, detection rate, and comprehensiveness, which might be greater diagnostic value in the joint fluid of infection after arthroplasty.

Introduction To assist orthopedic surgeons to select appropriate antibiotics, this study identified the common microorganisms that caused PJI and their drug-resistant spectrum. Methods. The clinical data of 202 patients with PJI after primary knee and hip arthroplasty from January 2017 to December 2021 were retrospectively analyzed. According to the time of postoperative infection, the patients were divided into acute PJI group (25 cases), delayed PJI group (91 cases), and chronic PJI group (86 cases). The results of pathogen species, composition ratio, drug susceptibility tests were collected. Results.The main pathogens of PJI were Gram-positive bacteria (58.91%, 119/202), and the positive rates of culture in patients with acute PJI, delayed PJI and chronic PJI were 32.00% (8/25), 62.64% (57/91) and 62.79% (54/86), followed by gram-negative bacteria (29/202, 14.36%) and fungus (4/202, 1.98%). Staphylococcus epidermidis and Staphylococcus aureus played an important role as well, followed by gram-negative bacteria (29/202, 14.36%) and fungus (4/202, 1.98%). Penicillin (81.25%), oxacillin (63.33%), erythromycin (61.17%) and clindamycin (48.35%) showed high antibiotic resistance rate in gram-positive bacteria. The drug-sensitivity to linezolid, vancomycin, daptomycin and tigecycline was 100%. The drug resistance rate to carbapenems was low, ranging from 0 to 3.57%. Conclusions. The main pathogens of PJI are Gram-positive bacteria, and the drug resistance rate of chronic PJI is higher than that of delayed and acute PJI.