Tiancai Liu

Hepatocellular carcinoma (HCC) is a malignant tumour with high global incidence and mortality, featuring limited therapeutic options and poor prognosis, which necessitates novel prognostic biomarkers. The immune checkpoint molecule CD274 (Programmed Death-Ligand 1, PD-L1) correlates with poor prognosis in most solid tumours, yet its prognostic value in HCC remains controversial, and circulating levels are unclear. This study aimed to multi-dimensionally investigate CD274's expression patterns and clinical significance in HCC. Bioinformatics analyses were performed on The cancer genome atlas (TCGA) and Genotype-Tissue Expression (GTEx) databases to screen differentially expressed genes, with functional and pathway annotations via Gene Ontology (GO), Kyoto Encyclopedia of Genes and Genomes (KEGG), Gene Set Enrichment Analysis (GSEA), and Single-Sample Gene Set Enrichment Analysis (ssGSEA). Survival analysis and univariate/multivariate Cox proportional hazards regression models were used to assess the prognostic value of CD274. PD-L1 protein expression was validated by immunohistochemistry (IHC) on HCC tissue microarrays. Serum PD-L1 concentrations were detected by ELISA in treatment-naive HCC patients and healthy controls. CD274 was lowly expressed in HCC tissues versus normal ones, and its expression correlated with TNF-α/NF-κB, complement, IFN-γ, adipogenesis and oxidative phosphorylation pathways, as well as lymphocyte-mediated immunity, immune globulin complexes and antigen binding. High CD274 expression negatively correlated with Th17 cells but positively with helper T cells, activated dendritic cells (aDCs), Th1 cells, T cells and macrophages via ssGSEA. Patients with high CD274 expression had longer overall survival (OS), disease-specific survival (DSS) and progression-free interval (PFI); multivariate Cox regression confirmed CD274 as an independent protective factor for DSS. PD-L1 protein expression showed no significant difference between HCC and normal liver tissues. Serum PD-L1 median was 349.6 pg/mL in HCC patients, significantly higher than 181.3 pg/mL in healthy controls. This study first identifies a unique "low tissue expression and elevated peripheral blood level" pattern of CD274 in HCC, with intratumoural high expression correlating with favourable prognosis. CD274 may exert distinct roles in local tumour and systemic immune regulation, and serum CD274 is a promising prognostic biomarker for HCC worthy of further research.

The integration of artificial intelligence (AI) and big data is poised to significantly augment drug research and development, offering the potential to address persistent challenges such as lengthy timelines and high failure rates. This review provides a critical overview of AI applications across the preclinical drug discovery pipeline (the 2020–2026 literature), covering drug–target interaction prediction, structure prediction, de novo design, virtual screening, drug repurposing, and ADMET forecasting. Beyond surveying technical developments, we critically discuss key translational hurdles, including data quality, model interpretability, patient heterogeneity, and regulatory adaptation, and provide structured summaries of representative models. We conclude by outlining future directions, such as multimodal AI, digital twins, and closed-loop automation, that aim to bridge the gap between computational prediction and clinical application. This review aims to inform researchers and accelerate the delivery of safe and effective therapies.

Cholangiocarcinoma (CCA) is a highly aggressive malignancy that is difficult to diagnose early and carries a poor prognosis. Conventional serological diagnostics are limited by inadequate sensitivity and the risks of invasive procedures, resulting in most patients being diagnosed at an advanced stage. In recent years, liquid biopsy has emerged as a novel approach for non-invasive and dynamic molecular tumor monitoring by detecting biomarkers such as circulating tumor cells (CTCs), extracellular vesicles (EVs), circulating tumor DNA (ctDNA) and clusterin (CLU). Furthermore, artificial intelligence (AI) has demonstrated strong potential in aiding diagnosis through medical image analysis, pathological pattern recognition, and multi-omics data integration, significantly improving the precision of early detection, risk stratification, and treatment response monitoring in CCA. This review systematically summarizes recent advances in liquid biopsy and AI for CCA diagnosis, discusses their clinical potential and current challenges, and offers perspectives on how their integration can propel the field toward earlier and more precise management of the disease.

Lung adenocarcinoma (LUAD) is a predominant form of lung cancer with poor prognosis due to early diagnosis challenges and treatment resistance. Thymidylate synthetase (TYMS), a key enzyme in DNA synthesis, has been associated with poor prognosis in multiple cancers, yet its multi‐omics characteristics in LUAD remain systematically uncharacterized. This study focuses specifically on LUAD, integrating transcriptomic, proteomic, and genomic data to elucidate TYMS expression patterns, prognostic value, immune microenvironment interactions, and therapeutic potential. Using data from The Cancer Genome Atlas (TCGA), the Clinical Proteomic Tumour Analysis Consortium (CPTAC), the Genome‐Tissue Expression Project (GTEx), the Gene Expression Profile Interaction Analysis (GEPIA2), etc., we conducted comprehensive bioinformatics analyses of TYMS in LUAD. TYMS showed significant overexpression at both mRNA and protein levels in LUAD tissue. High TYMS expression correlated with earlier tumour staging, higher histological grade, and poor prognosis. Functional enrichment analysis revealed TYMS involvement in p53/Rb and mammalian Target of Rapamycin (mTOR) signalling pathways. TYMS expression was associated with immune cell infiltration, showing negative correlation with B‐cell activity and positive correlation with neutrophil infiltration, as well as significant association with Cytotoxic T Lymphocyte‐Associated Protein 4 (CTLA4), Programmed Death‐1 (PDCD1, PD‐1), and Interleukin‐6 (IL‐6). Somatic copy number variation (SCNA) analysis indicated that TYMS amplification was linked to poor prognosis and enriched in the folate metabolism pathways. Drug sensitivity analysis demonstrated that high TYMS expression positively correlated with sensitivity to Afatinib and Gefitinib, but negatively correlated with Methotrexate and Vorinostat. In conclusion, TYMS is upregulated in LUAD and may serve as an independent prognostic biomarker and therapeutic target.

Tissue plasminogen activator - plasminogen activator inhibitor complex (tPAIc) is a critical biomarker to assess fibrinolytic dysfunction, which is widely used in clinics. Quality control material (QCM) plays an important role in immunoassays for human tPAIc. The QCM of tPAIc are derived from human plasma with many disadvantages. Recombinant protein is a promising substitute for human plasma to work as the source of QCM. However, tPAIc is a protein complex, consisting of three parts, tPA-A, tPA-B, and PAI-1, which makes the expression more difficult. This study aimed to obtain recombinant tPAIc QCM with excellent performance for immunoassay. Three recombinant plasmids that matched each part of the protein complex were constructed and co-transfected to HEK293F cells. The optimal molar ratio of three plasmids was further explored. Each part of the proteins was secreted from cells and the target protein tPAIc was self-assembled in the supernatant. After being identified by western blot and chemiluminescent immunoassay (CLIA), calculating the concentration of tPAIc in the supernatant, tPAIc was diluted to approximately 50 ng/mL and 5 ng/mL, distributed, and lyophilized in ampoules, working as QCM in tPAIc immunoassay. The homogeneity, stability, and recovery of the QCM were further evaluated. The three plasmids were successfully constructed. The target protein complex, tPAIc, was obtained in the supernatant at about 6500 ng/mL, under the best three plasmids co-transfected molar ratio 1:1:1. The QCMs were uniform in different ampoules. They were verified to be highly stable for at least 1 year when stored at 4 ℃ and − 20 ℃. The recombinant tPAIc QCMs for immunoassay were obtained with high quality to replace plasma-derived QCMs, which provides valuable insight into more application scenarios of recombinant proteins.

Scheme of the amplification-free digital droplet assay. Two innovations: (1) polystyrene capture beads to concentrate the target RNA from a dilute sample and (2) the “confinement” effect amplifies the concentration of the target molecule.