A molecular network analysis and in silico docking of beta-eudesmol, atractylodin and hinesol in patients with advance stage intrahepatic cholangiocarcinoma

ABSTRACT Cholangiocarcinoma (CCA), the bile duct cancer, is associated with a high burden and poor prognosis. This is due to the lack of early diagnostic tools and effective chemotherapy.

Molecular networking is a promising tool for investigating the molecular mechanisms of drugs or candidate molecules for various diseases. This study investigated molecular targets and

signaling pathways of the three components (atractylodin, beta-eudesmol, and hinesol) of _Atractylodes lancea_ Thunb. (DC.) (AL), the promising candidate for patients with advanced-stage

intrahepatic CCA (iCCA). The independent-sample T-test or Mann-Whitney U test was used to identify significant gene targets in (i) patients with advanced-stage iCCA who received AL treatment

and those who received palliative care alone, and (ii) patients with progressive and non-progressive diseases. A molecular network was constructed using Cytoscape to identify AL signaling

action pathways. Fifty-two genes were identified as the essential targeted genes in patients with advanced-stage iCCA. The most critical gene hubs were TNFα (1st rank), NRAS (2nd rank), and

PI3KCA (3rd rank). The false discovery rate (FDR) identified PI3K/AKT, NK cell-mediated cytotoxicity, and apoptosis as the top three significant pathways. Hinesol showed the highest binding

affinity compared with other components of AL and the standard anti-CCA drugs gemcitabine and 5-FU. Molecular networking is a valuable tool for investigating molecular signaling networks of

herbal medicine with multiple active and non-active ingredients. With multi-signaling targets linked to all tumor development and progression stages, the study supports AL as a promising

candidate for patients with advanced-stage iCCA. SIMILAR CONTENT BEING VIEWED BY OTHERS TANSHINONE IIA AFFECTS THE MALIGNANT GROWTH OF CHOLANGIOCARCINOMA CELLS BY INHIBITING THE

PI3K-AKT-MTOR PATHWAY Article Open access 29 September 2021 INVESTIGATING THE PI3K/AKT/MTOR AXIS IN BUZHONG YIQI DECOCTION’S ANTI-COLORECTAL CANCER ACTIVITY Article Open access 10 March 2025

NETWORK PHARMACOLOGY AND EXPERIMENTAL VERIFICATION BASED RESEARCH INTO THE EFFECT AND MECHANISM OF AUCKLANDIAE RADIX–AMOMI FRUCTUS AGAINST GASTRIC CANCER Article Open access 07 June 2022

INTRODUCTION Cholangiocarcinoma (CCA) is an aggressive malignancy often diagnosed in elderly patients who may not tolerate traditional chemotherapy due to adverse reactions and

comorbidities. This highlights the need for alternative treatments. Atractylodes lancea (Thunb.) DC (AL), a traditional herb used in East Asia, shows promise as a novel CCA therapy. Active

components like atractylodin and beta-eudesmol exhibit antiproliferative, anti-angiogenic, and anti-metastatic effects in preclinical studies1,2. Recent research highlights synergistic

interactions between AL’s bioactive constituents, such as beta-eudesmol, atractylodin, and hinesol, with favorable fractional inhibitory concentration (FIC) values3. These compounds induce

G1 cell cycle arrest and apoptosis in CCA cells by modulating p21 expression4. Beta-eudesmol also sensitizes chemotherapy-resistant CCA cells to agents like 5-FU and doxorubicin by altering

the Bax/Bcl-2 ratio and suppressing NQO1 expression5. It inhibits CCA cell migration by suppressing the epithelial-mesenchymal transition (EMT) pathway and downregulates the PI3K-AKT

signaling pathway6. In zebrafish models, beta-eudesmol shows anti-angiogenic effects by inhibiting Vegfa and Vegfr27. In vivo studies in a CCA hamster model further support these findings,

showing modulation of apoptosis-related genes8. Molecular networking, used in exploring drug and herbal interactions, has been applied in various diseases like rheumatoid arthritis9,

hepatocellular carcinoma10, and colorectal cancer11. While clinical trials have confirmed AL’s efficacy and safety in CCA treatment6, the detailed molecular mechanisms remain unclear. This

study aims to apply molecular networking to identify gene targets and signaling pathways modulated by AL in advanced-stage intrahepatic CCA (iCCA). It will focus on: (i) constructing a gene

ontology based on statistical analysis, (ii) identifying key gene targets through molecular network and functional enrichment analysis, and (iii) evaluating the binding affinity of AL’s

active compounds using in silico docking simulations. This approach will enhance understanding of AL’s mechanisms and its potential as a CCA therapeutic agent. METHODS STUDY DESIGN,

PATIENTS, AND STUDY PROCEDURES This study is part of a phase IIA open-label, randomized controlled trial assessing the dose, pharmacokinetics, safety, and efficacy of _Atractylodes

lancea_(AL) in patients with advanced-stage intrahepatic cholangiocarcinoma (iCCA)12 [01/03/2021; TCTR20210129007; ICTRP Search Portal (who.int)]. The protocol was approved by the Ethics

Committee of Sakol Nakorn Hospital (No. 049/2021) and conducted in compliance with the Helsinki Declaration. Informed consent was obtained from all participants. The AL treatment regimen

included: Group 1 (_n_ = 6): 1,000 mg daily of standardized AL extract for 90 days with standard care; 5 with non-progressive disease, 1 with progressive disease; Group 2 (_n_ = 5): Dose

escalation from 1,000 mg to 2,000 mg over 90 days with standard care; 3 with non-progressive disease, 2 with progressive disease; and Group 3 (_n_ = 4): Standard care alone; 3 with

progressive disease, 1 with non-progressive disease. Blood samples (3 mL) were collected on days 1 and 28. Plasma was isolated and stored at −80 °C for analysis. RNA was extracted from whole

blood, and gene expression was analyzed using the nCounter PanCancer Pathways Panel (NanoString Technologies, Seattle, US), which covered 730 genes linked to 13 cancer-related pathways. The

ratios of gene expression between day 28 and day 1 were analyzed to identify significant signaling pathways and differences between progressive and non-progressive disease, as well as

survivors and non-survivors. ANALYSIS OF THE ASSOCIATION BETWEEN GENE EXPRESSION AND AL TREATMENT Gene expression analysis in response to AL treatment involved a comprehensive network

analysis using functional enrichment through Gene Ontology (GO), Kyoto Encyclopedia of Genes and Genomes (KEGG) pathways, and Reactome. The comparisons included: (i) AL treatment (Groups 1

and 2) vs. non-treatment (Group 3), (ii) disease progression (progressive vs. non-progressive disease after AL treatment), and (iii) survival (survived vs. non-survived patients

post-treatment). Data normality was assessed using the Shapiro-Wilk test, with the independent-sample T-test for normally distributed variables and the Mann-Whitney U test for non-normally

distributed variables. The exact p-values were reported using SPSS version 23. Genes with significant expression changes were selected for molecular network analysis. The biological process

was the primary focus of molecular network analysis, as it captures the coordinated actions of molecular functions that collectively achieve biological objectives. This approach provides a

deeper understanding of the molecular mechanisms at play compared to examining individual molecular functions in isolation. Protein-protein interaction (PPI) networks were constructed using

Cytoscape, following these steps: (i) identification of significant gene expression variations between the AL-treated and control groups; (ii) access to the STRING database, selecting Homo

sapiens and full STRING networks with a confidence score cut-off of 0.4; (iii) importing gene expression data, including gene types, ratios, and p-values; (iv) functional enrichment analysis

through the cytoNCA plugin, which calculates betweenness, closeness, and degree centralities; and (v) iltering results to select KEGG pathways, Reactome pathways, and GO biological

processes. In the network, “degree” indicates a node’s direct connections, highlighting its role as a central hub. “Betweenness centrality” measures how often a node lies on the shortest

path between pairs of nodes, while “closeness centrality” indicates how efficiently a node connects to others in the network. GO and pathway enrichment analyses were performed using

Cytoscape 3.1.0 to visualize and annotate the results. Fisher’s exact test was used to select pathways involving at least three candidate genes, with statistical significance set at _p_ <

 0.05 and a false discovery rate (FDR) cut-off of _p_ < 0.05. The PPI network and pathway enrichment analysis leveraged STRING confidence scores and FDR values to identify significant

protein interactions and associated pathways. THE ASSOCIATIONS BETWEEN BIOMARKERS AND THE FOLD CHANGES OF GENE EXPRESSIONS IN PATIENTS WITH DISEASE PROGRESSION (PROGRESSIVE AND

NON-PROGRESSIVE DISEASES) OR SURVIVAL STATUS (SURVIVORS AND NON-SURVIVORS) Independent t-test (for normally distributed data) and Mann Whitney U test (for non-normally distributed data) were

used to compare the differences in biomarkers (fold changes in biomarker levels and mRNA expression) between the groups of patients with progressive and non-progressive diseases, as well as

between those who survived and did not survive. Pearson correlation test (for normally distributed data) and Spearman’s correlation test (for non-normally distributed data) were used to

determine the correlation between biomarkers (fold changes in biomarker levels and mRNA expression) in the groups of patients with progressive and non-progressive diseases, as well as in

those who survived and did not survive. The statistical significance level was set at α = 0.05 for all tests. ROC ANALYSIS, CUT-OFF VALUE, AND MODEL PERFORMANCE Receive Operating Curve (ROC)

was applied to determine the cut-off levels for significant levels of mRNA expressions between the groups of patients with progressive and non-progressive disease and those who survived and

did not survive. Area Under Curve (AUC) of ≥ 0.7 was the predefined criterion for identifying substantial mRNA expressions between progressive and non-progressive diseases and survivors and

non-survivors. A p-value of ≤ 0.05 was set as a statistically significant level. The Youden’s index was employed to determine the optimal cut-off value. Sensitivity, specificity, accuracy

index, Negative predictive value (NPV), and positive predictive value (PPV) were computed to indicate the model’s performance. KAPLAN-MEIER ANALYSIS The designated cut-off value for each

important gene was used to categorize the significant genes into two groups, i.e., ratios above or below the threshold. The Kaplan Meier assessed the impact of mRNA expression (disease

progression and survival status) above or below the threshold on progression-free survival time (PFS) and overall survival (OS). A log-rank test was applied to determine the significant

differences between ratios (above or below threshold) on PFS and OS with the predefined p-value of ≤ 0.05.Pathway cross-talk analysis. The Jaccard index was used to assess the reliability of

cross-talk analysis. The inclusion criterion was a pathway with _≥_2 candidate genes (to exclude genes with inadequate biological information) and the exclusion criterion was the pair of

pathways with < 2 shared candidate genes. The selected pair of pathways were prioritized based on the Jaccard index of > 0.25 and > 0.6 for pathway-pathway interaction and pathway

redundancy, respectively13. IN SILICO DOCKING SIMULATION OF GENE TARGETS In silico docking simulations were conducted using AutoDock software to evaluate the binding affinities of

Atractylodes lancea (AL) compounds (atractylodin, beta-eudesmol, and hinesol), standard anti-CCA drugs (5-FU and gemcitabine), and reference compounds to ten significant target proteins:

JAK1, PI3KCA, HDAC2, ILR2A, ILR2B, TNF-α, SMAD2, NRAS, SOS1, ERRB2, and DMT3A. The crystallographic structures of the AL target proteins were retrieved from the RCSB Protein Data Bank, and

the active compound structures were obtained from the PubChem database in SDF format, then converted to PDB files using PyMOL or OpenBabel. A free energy force field was calculated using a

Lamarckian Genetic Algorithm to predict the binding affinity based on the Gibbs free energy. Docking was performed with the catalytic domain of each receptor to interact with the small

ligands (Table S1). The optimal docking conformation was selected based on the ligand conformation with the lowest root-mean-square deviation (RMSD) using a tolerance of 2.0 Å. A binding

affinity of ≥ 6 kcal/mol was considered acceptable. 5-FU and gemcitabine were chosen for docking simulations as they are first-line treatments for intrahepatic cholangiocarcinoma (iCCA).

RESULTS BASELINE CHARACTERISTICS The diagrammatic workflow in this study is shown in Fig. 1. Table 1 summarizes the baseline characteristics of the patients with advanced-stage iCCA in each

treatment group (group 1 and group 2) and the control group (group 3). No significant differences between AL-treated patients with progressive and non-progressive diseases were observed with

age (_p_ = 0.843), body weight (_p_ = 0.278), height (_p_ = 0.768), white blood cell count (_p_ = 0.212), platelet count (_p_ = 0.125), hemoglobin (_p_ = 0.105), INR (_p_ = 0.914), PT (_p_ 

= 0.753), PTT (_p_ = 0.694), creatinine clearance (_p_ = 0.471), AST (_p_ = 0.053), ALT (_p_ = 0.078), ALP (_p_ = 0.809), CPK (_p_ = 0.103), LDH (_p_ = 1.00), CA19-9 (_p_ = 0.661), CEA (_p_ 

= 0.177), and IL-6 (_p_ = 0.343) between the treatment and control groups. Subgroup analysis revealed no significant changes in baseline AST (_p_ = 0.842), ALT (_p_ = 0.630), ALP (_p_ = 

0.352), LDH (_p_ = 0.376), CPK (_p_ = 0.09), CA19-9 (_p_ = 0.497), CEA (_p_ = 1), and IL6 (_p_ = 0.683) levels. The comparable values between survivors and non-survivors in the AL-treated

group were not substantially different [(AST, _p_ = 0.145), (ALT, _p_ = 0.352), (ALP, _p_ = 0.886), (LDH, _p_ = 1.00), (CPK, _p_ = 0.327), (CA19-9, _p_ = 1.00), (CEA, _p_ = 0.727), and (IL6,

_p_ = 0.183). ASSOCIATION BETWEEN GENE EXPRESSION AND AL TREATMENT Out of 730 genes analyzed, 52 showed significant changes in gene expression levels following AL treatment (Fig. 2).

PRKAR2A, SPOP, RXRG, FGF1, PAX8, TNFRSF10A, and TCFL1 genes showed high gene expression levels, while the remaining showed relatively low expression levels. Further analysis revealed a

marked difference in the expression levels of 16 gene targets in patients with progressive and non-progressive disease following AL therapy (Group 1 and 2). BDNF, ERBB2, TTK, and XPA showed

higher expression levels, while the remaining showed lower expression levels. Seven genes were identified with marked differences between survivors and non-survivors. HSPA1A, FOS, CBL, and

MMP9 were the top four upregulated genes. Tables S2, S3 and S4 summarize significant gene expressions (fold-change) between AL-treated and control group, and patients with progressive and

non-progressive diseases, and survivors and non-survivors, respectively. PROPOSED MOLECULAR SIGNALING PATHWAYS OF AL ACTION IN ADVANCED-STAGE ICCA A comprehensive network analysis in

AL-treated (Groups 1 and 2) and control (Group 3) identified 44 potential target genes. Thirty genes were identified as core hubs (Table 2). TNF-α was the network’s most significant hub (PPI

 = 22), followed by PIK3CA (PPI = 15), NRAS (PPI = 14), JAK1 (PPI = 11), IL2RB (PPI = 11), and IL2RA (PPI = 11). Subsequent network interaction analysis revealed 44 nodes and 115 edges (PPI 

= 5.18). Based on pathway enrichment analysis, AL action (specific condition or disease) was significantly linked with various signaling pathways, i.e., PI3K-AKT, NK cell-mediated

cytotoxicity, cellular apoptosis, MAPK, JAK-STAT and NF-kB in cancer, EGFR-resistance, focal adhesion, Fc epsilon RI, T cell receptor, cell senescence, regulation of actin cytoskeleton,

transcription misregulation, B cell receptor, mTOR, ERBB, Ras, VEGFR, PDL-1, RIG-I-like receptor, platinum drug resistance, and P53 signaling pathways (Fig. 3). Comprehensive network

analysis using Reactome database identified 38 multiple signaling pathways, i.e., signal transduction, IL2, IL3, IL4, IL5, IL13, GM-CSF, ILR SHC, extra-nuclear estrogen, FGFR3 in disease,

downstream signaling of activated FGFR1-4, DAP12, death receptor, FLT3, PDGFRA, and EGFR-v-III. These pathways were linked with 180 biochemical pathways. Functional enrichment analysis of

the 20 selected biological processes based on the set criterion, i.e., KEGG, biological processes, and Reactome are shown in Figs. 3 and 4, and 5, respectively. The set criterion included

the three shared genes with a p-value of ≤ 0.05 between signaling pathways for representing the KEGG, biological, and Reactome processes. A pharmacological mechanism of AL is shown in Fig. 

6. THE ASSOCIATIONS BETWEEN BIOMARKERS AND THE FOLD CHANGES OF MRNA EXPRESSIONS IN PATIENTS WITH DISEASE PROGRESSION AND SURVIVAL STATUS, KAPLAN-MEIER ANALYSIS, AND COX-PROPORTIONAL

REGRESSION ANALYSIS DISEASE PROGRESSION The study assessed the relationship between proposed biomarkers and mRNA expression changes in patients experiencing disease progression and survival

status through Kaplan-Meier and Cox-proportional regression analyses. A significant difference in ALP ratios was observed between patients with progressive and non-progressive disease, with

a p-value of 0.048. Other biomarkers (AST, ALT, LDH, CPK, CA19-9, CEA, and IL-6) did not show significant associations (Table S5). The ROC analysis indicated that the cut-off value for ALP

ratios to distinguish between disease progression was ≥ 1.023 folds (AUC = 0.917, 95% CI = 0.739–1.00, _p_ = 0.04; Youden’s index = 0.875) (Fig S1). This yielded a sensitivity of 87.5%,

specificity of 100%, positive predictive value (PPV) of 1, negative predictive value (NPV) of 0.75, and accuracy index (AI) of 0.9. However, Kaplan-Meier analysis revealed significant no

differences in progression-free survival (PFS) based on these ALP ratios (_p_ = 0.104) (Fig S2). No significant difference following cox-proportional regression analysis was found disease

progression and ALP ratios (HR: 5.27, 95%CI: 0.89–100.6, _p_ = 0.1265) (high vs. low) (Fig S3). The ROC analysis for various genes indicated significant differences in their expression

between progressive and non-progressive disease: APC (≥ 1.49 folds, AUC = 0.917, _p_ < 0.001), BAX (≥ 0.82 folds, AUC = 0.917, _p_ < 0.001), DNMT3A (≥ 0.805 folds, AUC = 1, _p_ < 

0.001). Others included ETS2, GNA11, IGF1, IRAK2, PLCG2, POLB, RAF1, THEM4, BDNF, CDKN2A, ERBB2, TTK, and XPA, with respective AUC values indicating strong predictive capability (AUCs ≥ 

0.917) (Fig S4 ,5). Significant associations with PFS were noted for DMNT3A, GNA11, IGF1, IRAK2, POLB, RAF1, THEM4, and XPA (Table S6, Fig S6-21). RAF1 and XPA demonstrated significant

hazard ratios 9.316 (_p_ = 0.042) and 0.174 (Reciprocal HR: 5.732) (_p_ = 0.037) (Fig S22-36, Table S7), respectively, indicating a strong association with disease progression. A correlation

analysis revealed positive correlations between ALP ratios and BAX, PLCG2, and RAF1, while BDNF and TTK showed negative correlations (Table S8). Notably, in patients with non-progressive

disease, a strong positive correlation (_r_ = 0.998, _p_ = 0.037) between ALP and GNA11 ratios was found, alongside a significant negative correlation with TTK (_r_=−1.00, _p_ = 0.009) in

patients responding to AL therapy (Table S9). SURVIVAL STATUS No significant differences in known biomarkers were observed between survivors and non-survivors receiving AL therapy (p-values

ranged from 0.145 to 1.00) (Table S10). ROC analysis revealed significant differences for AR and EYA1 ratios, both demonstrating AUC values of 1 (_p_ = 0.034), with cut-off values of ≥ 1.895

and ≥ 1.67 folds, respectively (Fig S37, Table S11). Significant differences in overall survival (OS) were observed for AR and EYA1 (_p_ = 0.008) (Fig S38, 39, Table S12). The Cox hazard

ratios did not demonstrate significant predictive value (_p_ = 0.504) (Fig S40, 41, Table S13). Strong positive correlations were found between AR and CA19-9 ratios (Table S14), and between

EYA1 and CA19-9 ratios (Table S15), indicating a potential association of these biomarkers with survival outcomes. In addition, negative correlations were reported between AR and ALT ratios

(Table S14), and between EYA1 and CPK ratios (Table S15, indicating an association of these biomarkers with survival outcomes. Logistic regression highlighted RAF1 and XPA as significant

predictors for disease progression, with acceptable McFadden’s scores indicating model reliability. Multi-logistic regression suggested that AR and EYA1 were significantly associated with

OS, but the high McFadden’s score indicated potential model inadequacy. Probability risks for disease progression were stratified into four levels based on RAF1 and XPA ratios, suggesting

tailored approaches for patient management based on risk assessment. ADVERSE DRUG REACTIONS The incidence rates of grade 2 and 3 adverse events were consistent across groups, with no

significant differences in blood parameters or liver function tests, suggesting minimal AL-related adverse effects. CROSS-TALK AMONG POTENTIAL TARGET GENES AND SIGNALING PATHWAYS Twenty-four

enriched pathways were investigated for the interplay between the two functional interaction pathways. The cross-talk signaling pathways associated with tumor growth/proliferation/ survival

of cancer, angiogenesis, and metastasis were categorized into three groups according to their abilities to depict the mechanisms involved in tumor formation and invasion. The major

cross-talk module consisted of various cellular processes, including cellular response to external stimuli, cell communication, regulation of cell death, signal transduction, and immune

response. Table 3 summarizes the cross-talk signaling pathways associated with tumor growth, proliferation, progression, survival, angiogenesis, and metastasis in patients with

advanced-stage iCCA following AL therapy. _Tumor growth/proliferation/survival pathway_ The PI3K-AKT and JAK-STAT pathways demonstrated the highest similarity in tumor growth, proliferation,

and survival, with a Jaccard value of 0.54, indicating significant overlap in the shared genes between the two pathways. The second-highest ranked pathways were MAPK-T cell receptor,

NK-cell mediated toxicity-B cell receptor, Fc epsilon RI-proteoglycan, proteoglycan-T cell receptors, proteoglycan-NK-cell mediated cytotoxicity, PDL-1-Ras, PDL-1, and TNF-α signaling

pathways. _Angiogenesis_ The highest-ranked pathways for angiogenesis included EGFR resistance-mTOR, EGFR resistance-Ras, EGFR resistance-PDL1, EGFR resistance-ErBb, ErBb-PDL1, ErBb-Ras, and

ErBb-proteoglycan signaling, all showing the same level of similarity with a Jaccard index of 0.5, indicating significant cross-linking in their involvement. _Metastasis_ The regulation of

actin cytoskeleton-Fc epsilon RI pathway had the highest Jaccard index for metastasis. CANDIDATE GENE TARGETS OF AL GENE TARGETS ASSOCIATED WITH DISEASE PROGRESSION Eleven gene targets and

18 edges, with an average degree of 3.27 [PPI], were used to build a comprehensive pharmacology-based network analysis in patients with progressive compared with non-progressive diseases.

ERBB2 (1st rank), CDKN2A (2nd rank), BDNF (3rd rank), and IGF-1 (4th rank), DNMT3A (4th rank), Raf (4th rank) and GNA11 (4th rank) showed degrees of 8, 6, 4, 3, 3, 3, and 3, respectively,

indicating their potential central hubs in regulating iCCA progression. Three different analyses for regulating iCCA progression following functional analysis based on the eleven gene

targets including KEGG pathway, biological processes, and Reactome pathway, were determined as follows: i) KEGG analysis of the modulatory effects of AL on various signaling pathways such as

the EGFR tyrosine kinase inhibitor, micro-RNA in cancer, Ras, TP53, platinum-drug resistance, MAPK, PI3K-AKT, ERBB2, and HIF-signaling pathways, II) signaling transduction of various

biological processes, e.g., regulation of the cellular metabolic process and protein phosphorylation, and intracellular signal transduction, III) Reactome analysis of platelet activation and

aggregation signaling pathways associated with GNA11, RAF1, IGF1, and PLCG2 target genes based on protein-protein interaction analysis. GENE TARGETS ASSOCIATED WITH SURVIVAL A comprehensive

pharmacology-based network analysis was performed in survivors and non-survivors using seven gene targets and five edges, with an average degree [PPI] of 1.43. AR, FOS, HSPA1A, and MMP9 had

degrees of 3 (1st rank), 3 (1st rank), 2 (2nd rank), and 2 (2nd rank). Functional enrichment analysis suggests that AL modulates toll-like receptor cascades, IL12 family signaling,

regulation of mRNA stability, autophagy, signaling by NOTCH, regulation of TP53 activity, signaling by TGF-beta receptor complex, and MAPK family signaling cascades. BINDING AFFINITY OF AL

TO CANDIDATE GENE TARGETS Table 4 summarizes the binding affinities of the three components of AL (atractylodin, beta-eudesmol, and atractylodin), standard drugs for iCCA (gemcitabine and

5-FU), and reference drugs/compounds (Ruxolitinib, idelalisib, panobinostat, SM16, GTP, BI-3406, Afatinib, and guadecitabine) to eleven candidate molecular targets of AL in iCCA. All

reference drugs showed the highest binding affinities to all targets. Among the three components of AL, hinesol showed the highest binding affinities to all targets. The binding affinity of

beta-eudesmol was relatively lower. Atractylodin weakly binds to most targets. For standard drugs, gemcitabine showed a high binding affinity comparable to hinesol. Residue proteins and root

mean square deviation (RMSD) for each drug and target (binding affinity ≥ 6 kcal mol−1) are shown in Supplementary figures S42-77. DISCUSSION ASSOCIATIONS BETWEEN BIOMARKERS AND DISEASE

PROGRESSION This study identified an ALP ratio increase (≥ 1.024) as a potential predictor of iCCA progression following AL treatment. However, Kaplan-Meier analysis showed no significant

correlation between ALP changes and progression-free survival (PFS), suggesting ALP ratios alone do not independently predict treatment response. Elevated ALP levels are linked to various

hepatobiliary and metastatic conditions, including CCA, where high ALP correlates with rapid progression after gemcitabine-cisplatin treatment (_p_= 0.01)14. Additionally, high platelet

counts and lower hemoglobin levels were associated with a worse prognosis, with hemoglobin showing a hazard ratio (HR) of 0.90 (95% CI = 0.83–0.97, _p_= 0.009)15. Regarding mRNA expression,

11 of 16 genes demonstrated predictive potential for AL response, with an area under the curve (AUC) > 0.9 (_p_ < 0.001). Patients with RAF1 ratios ≥ 0.5 had a 9.32-fold higher risk of

disease progression, while those with XPA levels < 1.34 had a 5.73-fold increased risk, emphasizing their role as early prognostic markers for AL therapy continuation. Elevated RAF1

ratios indicate heightened progression risk, whereas lower XPA levels suggest impaired DNA repair, contributing to disease advancement. Unlike targeted mRNA/protein-based therapies,

conventional chemotherapy (gemcitabine/cisplatin) lacks specific molecular targets. However, overexpression of ATP-binding cassette (ABC) transporters, such as ABCC1 (MRP1), reduces

intracellular drug concentrations, contributing to 5-FU resistance16. Similarly, altered orotate phosphoribosyl transferase activity weakens FdUMP activation, impairing 5-FU efficacy, while

OCT4 and MRP1 upregulation have been implicated in gemcitabine resistance17. Furthermore, FGFR2 alterations impact AKT/mTOR and STAT3 pathways, linking them to gemcitabine resistance,

reinforcing the need for FGFR2 testing before treatment18. For targeted therapy, risk-stratified gene expression analysis identified HACL1, LAMA4, GMNN, C1RL, PCBD1, and FXYD2 as prognostic

markers. The low-risk group responded better to bosutinib, gefitinib, gemcitabine, and paclitaxel, whereas the high-risk group showed improved responses to axitinib, cisplatin, and

imatinib19. Additionally, PD-L1 expression serves as a prognostic indicator for immune checkpoint inhibitor (ICI) efficacy in iCCA patients. Notably, there was no overlap in the genetic

markers associated with AL treatment and those linked to conventional or targeted therapies, suggesting distinct molecular pathways drive progressive disease under different therapeutic

regimens. CORRELATION ANALYSIS AND PROGNOSTIC MARKERS PLCG2 showed the strongest positive correlation with ALP (_r_ = 0.736, _p_ = 0.01), linking its activation of the ERK pathway to iCCA

proliferation and apoptotic resistance. Patients with elevated PLCG2 levels experienced worse disease progression (1.40 vs. 0.41). Conversely, TTK, a promoter of tumor survival in gastric

and hepatocellular cancers, exhibited a moderate inverse correlation with ALP (_r_=−0.727, _p_ = 0.01). These findings suggest that genetic signatures play a role in cancer biomarker

expression and disease progression. RAF1 ratios correlated with increased progression risk, while lower XPA ratios indicated reduced risk. Raf-1 inhibition has been shown to suppress CCA

metastasis by modulating MMP9/MMP4 expression and inhibiting EGF-induced cell growth20. Clinical studies reported an 18.1% partial response rate for Ras inhibitors across multiple cancers,

including CCA21. Re-challenging Raf-1 inhibitors in BRAF-1 aberrant patients also yielded clinical responses, supporting its therapeutic potential. XPA, crucial for nucleotide excision

repair (NER), mediates platinum-based chemotherapy resistance22. In progressive iCCA cases, mRNA expression of XPA was elevated; however, XPA ratios < 1 suggest AL may suppress its

expression, potentially reversing platinum resistance. Combining AL with gemcitabine-cisplatin may enhance disease control rates (DCR) in patients with platinum-resistant iCCA compared to

chemotherapy alone. SURVIVAL ANALYSIS This study found no significant differences in liver enzymes, CA19-9, CEA, or IL-6 between AL-treated iCCA patients and controls, suggesting these

biomarkers may not predict survival in this cohort. However, the small sample size, with only two deaths recorded, may limit statistical power. Previous studies have shown that ALP levels ≥ 

138 IU/L were associated with shorter overall survival (HR = 1.654, 95% CI = 1.170–2.38, _p_= 0.004) following chemotherapy, compared to ALP < 138 IU/L (11.6 vs. 4.47 months)23.

Similarly, CA19-9 < 1,000 IU/mL correlated with improved one-year survival23. While both gemcitabine and 5-FU influenced ALP levels, only gemcitabine-based treatment showed significant

changes23. Jiang et al. reported that iCCA patients with ALP ≤ 147 IU/L had longer one-year and three-year survival (23 vs. 10.4 months) and a 1.64-fold higher probability of survival (95%

CI = 1.20–2.22, _p_= 0.002) than those with ALP > 147 IU/L24. Higher CA19-9 levels were linked to better one-year survival than ALP ≤ 37 µg/mL (29.2 vs. 14.1 months), with a 1.80-fold

increased survival probability (95% CI = 1.337–2.445, _p_< 0.001) in those with ALP ≤ 37 µg/mL25. These findings highlight the prognostic value of ALP and CA19-9 in iCCA, though their

relevance in AL-treated patients remains unclear. PROMISING CANDIDATE GENE TARGETS FOR AL AL therapy targets critical pathways involved in iCCA progression, including PI3K-AKT, JAK-STAT, and

Ras/MAPK/mTOR, which regulate tumor growth, survival, and immune evasion26,27,28,29,30,31,32,33,34. AL enhances immune response by stimulating NK, T-, and B-cell activity through genes such

as GZMB, TNFRSF10A, PRKCG, PIK3CA, and SYK, which are linked to NK cell-mediated cytotoxicity and adaptive immunity35,36,37,38,39,40. Downregulation of tumor-promoting genes, including

GNA11 (3.5-fold), IGF1 (7.2-fold), THEM4 (13.15-fold), PLCG2 (8.2-fold), POLB (7.14-fold), and RAF1 (17.35-fold), correlates with reduced tumor proliferation, differentiation, invasion, and

metastasis41,42,43,44,45. Suppressing RAF1 may limit EGFR resistance and enhance the efficacy of EGFR/PI3K inhibitors, potentially overcoming drug resistance in iCCA46,47,48,49,50,51.

Additionally, suppression of TSLP (7-fold), COL5A2, FLNC, ITGA3, and PLAU may improve overall survival by reducing tumor progression, angiogenesis, and immune suppression52,53,54,55,56,57.

AL therapy appears to shift gene expression toward tumor suppression, supporting its potential as an adjunct treatment for advanced iCCA by inhibiting proliferation, reversing resistance,

and enhancing immune response. PROPOSED MOLECULAR TARGETS OF AL ACTION KEGG pathway enrichment analysis identified six critical pathways involved in iCCA progression following AL therapy:

tumor proliferation, growth, survival, angiogenesis, invasion, and metastasis. The PI3K/AKT pathway (FDR = 2.26E-23) was the most significantly altered, impacting tumor progression through

mTOR, Raf-1, NF-κB, MDM2, PDK1, p21/p27, Caspase 3/7/9, FOXO, eNOS, and

MMP-2/9/1317,18,34,41,42,43,44,45,53,54,55,56,57,58,59,60,61,62,63,64,65,66,67,68,69,70,71,72,73,74,75,76,77,78,79,80,81,82. AL therapy modulated 15 key genes linked to the PI3K/AKT pathway,

including OSM, SOS1, IL2RA, IL2RB, NRAS, NOS3, EIF4EBP1, JAK1, SYK, BCL-2, FGF1, IKBKG, ITGA3, GNGT1, and PIK3CA. Notably, JAK1 (33-fold downregulated) regulates tumor growth and immune

response, and its suppression may control iCCA progression and reduce immunosuppressive effects17,31,54,70,71,72,73,74. OSM (downregulated), associated with lymphatic and distant metastasis,

correlates with better overall survival (OS) and recurrence-free survival (RFS)54. NOS3 (downregulated) suppresses apoptosis and promotes angiogenesis via VEGF; its downregulation may

inhibit metastasis43,76. EIF4EBP1 (downregulated) is linked to poor differentiation and lymph node metastasis in hilar CCA, and suppression may reduce tumor invasion28. SOS1 (22-fold

downregulated) regulates PI3K/AKT, EMT, and MAPK signaling, and its suppression could inhibit EGFR resistance, EMT, and metastasis78,79,80. ITGA3 (8-fold downregulated) activates PI3K

through focal adhesion; downregulation correlates with reduced tumor size, lymph node metastasis, and advanced TNM stage56. BCL-2 (7-fold downregulated) is linked to chemoresistance, and its

suppression may enhance apoptosis and treatment response46,75. FGF1 (14-fold upregulated) potentially activates angiogenesis via PI3K/AKT, although its impact may be minimal due to low

FGFR1 overexpression in iCCA83,84,85. Downregulation of key oncogenic genes in the PI3K/AKT pathway suggests that AL therapy suppresses iCCA proliferation, growth, angiogenesis, and

metastasis. The 6-fold downregulation of PIK3CA indicates a significant reduction in pathway activation, supporting the observed improvements in OS and progression-free survival (PFS) in

AL-treated patients compared to palliative care. These findings position AL as a promising adjunct therapy in advanced-stage iCCA. NK CELL-MEDIATED CYTOTOXICITY The NK cell-mediated

cytotoxicity pathway was the second most altered pathway following AL therapy, involving key immune response regulators like GZMB, TNFRSF10A, TNF-α, PKCG, and PIK3CA84. GZMB (downregulated)

is essential for apoptosis via caspase-3/7 activation, and its reduction may impair NK and T-cell cytotoxicity, affecting immunosuppressive functions52. TNFRSF10A (10-fold upregulated), also

known as TRAIL-1, promotes apoptosis in iCCA cells through DR4/5 activation. Its increase may compensate for GZMB downregulation, enhancing tumor cell death29. TNF-α (downregulated) plays a

role in NK cell proliferation and cytotoxicity, but its reduction did not significantly affect CDKN2A expression, indicating stable NK cell function80,81. AL therapy modulates NK

cell-mediated immune responses, potentially suppressing granzyme-mediated apoptosis while enhancing TRAIL-dependent apoptosis, supporting its anti-tumor effects in iCCA. Additionally,

downregulation of key oncogenic genes in the PI3K/AKT pathway suggests AL therapy inhibits iCCA proliferation, growth, angiogenesis, and metastasis. Notably, the 6-fold downregulation of

PIK3CA indicates a significant reduction in pathway activation, supporting the observed improvements in OS and PFS in AL-treated patients compared to palliative care. These findings position

AL as a promising adjunct therapy for advanced-stage iCCA. APOPTOSIS PATHWAY The apoptosis pathway ranked third among the pathways linked to AL action, involving genes such as GZMB,

TNFRSF10A, PI3KCA, BCL-2A1, MAP3K5, NRAS, BCL-2, TNF, and IKBKG. BCL-2A1, an anti-apoptotic protein, promotes survival in various cancers, with higher overexpression in lymphoma and leukemia

patients compared to those with solid tumors86. In advanced-stage solid tumors, BCL-2A1 upregulation is linked to chemoresistance86. The downregulation of both BCL-2A1 and BCL-2 after AL

therapy may increase sensitivity to chemotherapy-resistant iCCA. IN SILICO DOCKING SIMULATION In silico docking results revealed high binding affinities of hinesol and beta-eudesmol to the

candidate molecular targets, supporting AL’s clinical efficacy in a phase IIA trial12. While hinesol showed low cytotoxic potency against iCCA cells87, it had high binding affinity to all

targets except HDAC2, NRAS, and SOS1. Notably, hinesol combined with beta-eudesmol or atractylodin exhibited synergistic cytotoxic effects across all iCCA cell lines87. These findings

suggest that the compounds effectively bind to their molecular targets, preventing tumor development and inducing cell death. Gemcitabine demonstrated higher binding affinity than 5-FU to

all targets, likely due to the differing molecular targets of the two drugs—5-FU targets thymidylate synthase, while gemcitabine targets DNA replication components88. Patients treated with

gemcitabine plus cisplatin89showed superior tumor control, time-to-progression, disease-free survival, and overall survival compared to those treated with 5-FU, leucovorin, and cisplatin89.

5-FU and gemcitabine were selected for docking simulation as first-line treatments for intrahepatic cholangiocarcinoma, although their mechanisms of action are limited to targeting

AL-related molecules such as JAK1, PI3KCA, HDAC2, and IL2RA. Further preclinical studies are needed to confirm the efficacy of gemcitabine and 5-FU on these potential targets. POTENTIAL

ERRORS AND BIAS Potential errors in ligand-receptor interactions affecting ligand binding affinity can be categorized into four sources as follows: _Protein structure_ Missing residues or

domains may alter the binding pocket geometry. Incorrect rotamer assignments within the binding pocket can misalign interactions with the ligand. Protein flexibility is often overlooked.

_Ligand_ Inadequate sampling of ligand conformations may underestimate the optimal binding pose. Incorrect protonation states of functional groups can affect receptor interactions.

Inaccurate force field parameters can lead to incorrect energy calculations. _Docking algorithm_ Search algorithms may fail to sufficiently sample potential binding modes, especially in

complex interactions. Improperly defined binding sites may lead to irrelevant biological targets. _Environmental factors_ Not accounting for water molecules around the binding site can

distort protein-ligand interaction. EVIDENCE BEFORE THIS STUDY Beta-eudesmol, atractylodin, and hinesol --key active compounds of _Atractylodes lancea_(Thunb) DC exhibit antiproliferative,

apoptotic, anti-metastatic, and anti-angiogenic effects through multiple mechanisms: (i) promotion of cell cycle arrest through modulation of p16, CDK1, CDK4, and p218; (ii) apoptosis

activation through increasing Bax/Bcl2 ratio, TNFRSF6, cytochrome C, and caspase activation (caspase-3, −8, −9, Apaf1, Bax-3)8,90; (iii) immunostimulation through enhancing CDKN2B

expression, promoting immune control of iCCA90; (iv) PI3K-AKT pathway suppression through downregulating phosphorylated PI3K, AKT, p-p38MAPK, STAT1/3, HO1, and NF-κB6,91; (v)

anti-angiogenesis through downregulating Vegfa and Vegfr2 expression7; and (vi) anti-metastatic effects through downregulating MMP9, N-Cadherin, and TGF-beta expression8. These findings stem

from in vitro and in vivo studies; however, clinical validation in humans remains lacking. ADDED VALUE OF THIS STUDY This is the first study to provide a network analysis of AL and its

bioactive constituents—beta-eudesmol, atractylodin, and hinesol—in patients with advanced intrahepatic cholangiocarcinoma (iCCA). Our findings demonstrate that AL inhibits iCCA growth

through multiple mechanisms. AL promotes apoptosis via Bcl-2, Bcl-2A1, ASK, and SMAD2, enhances NK cell-mediated cytotoxicity through FAS-TRAIL, TNFRSF10A, PKCG, and TNF-alpha, and

suppresses CCA metastasis by targeting ITGA3 and EIF4EBP1. It also affects angiogenesis by interacting with NOS3, ERBB2, and FGF1/FGFR1. Additionally, AL sensitizes CCA to

gemcitabine/5-FU-induced chemoresistance by modulating the Bcl-2/Bcl-2A1 pathway. Docking simulations confirm that AL bioactives bind to key targets, including JAK-1, HDAC2, IL2RA, SMAD2,

NRAS, SOS1, ERBB2, and DNMT3A, which are involved in metastasis and angiogenesis. Notably, RAF1 and XPA were identified as potential new targets for CCA therapies. IMPLICATIONS OF ALL THE

AVAILABLE EVIDENCE Our data, consistent with previous studies, show that AL and its bioactive compounds significantly control cholangiocarcinoma (CCA) through multiple mechanisms, including

enhancing apoptosis, inhibiting cellular growth, boosting NK cell-mediated cytotoxicity, preventing angiogenesis, and inhibiting invasion. We also identify new targeted proteins for CCA

therapy. Previous research highlights AL’s antiproliferative effects, mainly via the PI3K/AKT pathway, and its role in caspase-dependent and independent apoptosis, anti-angiogenesis, and

tumor invasion prevention. Through network analysis, we identified new targets and confirmed existing ones. Notably, patients treated with AL and exhibiting non-progressive disease (with

lower RAF1 and higher XPA expressions) provide valuable insights for developing new drugs and treatment strategies for intrahepatic cholangiocarcinoma (iCCA). IMPACTS OF THE STUDY Combining

AL with a first-line regimen (gemcitabine plus cisplatin) may improve the disease control rate (DCR) in patients with platinum-based chemotherapy resistance. This study’s results, using a

probability approach, serve as early prognostic markers, aiding in predicting the DCR, minimizing unnecessary treatments, and maximizing efficacy in clinical settings. LIMITATIONS The

statistical analysis of multiple comparisons may be limited by a false positive rate, as Bonferroni correction was not applied. Additionally, the molecular network relies on static

protein-protein interactions (PPI), while PPIs are dynamic, and gene expression can change over time. mRNA expression may not always correlate with protein translation, and further

validation through protein expression, proteomics, and functional analysis, including western blot and kinetic binding assays, is needed to confirm these results. While docking simulations

provide useful insights into protein-ligand binding affinity, their accuracy is influenced by the quality of input data (protein structure, ligand quality, docking algorithm, and

environmental factors). Neglecting protein flexibility, as seen with AutoDock Vina, may lead to inaccurate binding affinity calculations. Further assays, including western blot analysis and

kinetic binding studies, are necessary to validate the interactions between AL bioactives and their targets. The Lamarckian Genetic Algorithm (LGA) in AutoDock accounts for flexible ligand

poses, helping predict binding affinity for flexible proteins in biological contexts. However, limitations like atom type constraints and memory issues can result in long computational

times. These limitations can be minimized by using well-refined protein structures (from X-ray crystallography or NMR with ≤ 3 Å resolution), selecting complete protein residues, defining

binding pockets based on existing literature, and considering the role of surrounding water molecules. CONCLUSION A molecular networking study confirmed AL as a potential alternative

treatment for iCCA by targeting key pathways such as PI3K/AKT, NK cell-mediated cytotoxicity, and apoptosis. These pathways regulate iCCA development, progression, angiogenesis, and

invasion. This study highlights the utility of molecular networking in identifying signaling targets of herbal medicines, which contain both active and non-active components. DATA

AVAILABILITY The authors confirm that the data supporting the findings of this study are available within the article and/or its Supplementary Materials and Methods. The datasets used and/or

analysed during the current study are available from the corresponding author upon reasonable request. REFERENCES * Banales, J. M. et al. Cholangiocarcinoma 2020: The next horizon in

mechanisms and management. _Nat. Rev. Gastroenterol. Hepatol._ 17:9, 557–588. https://doi.org/10.1038/s41575-020-0310-z (2020). Epub 2020 Jun 30. Article  Google Scholar  * Na-Bangchang, K.,

Plengsuriyakarn, T. & Karbwang, J. Research and development of atractylodes lancea (Thunb) DC. as a promising candidate for cholangiocarcinoma chemotherapeutics. _Evid. Based

Complement. Alternat Med._ 2017, 5929234. https://doi.org/10.1155/2017/5929234 (2017). Article  PubMed  PubMed Central  Google Scholar  * Martviset, P., Chaijaroenkul, W., Muhamad, P. &

Na-Bangchang, K. Bioactive constituents isolated from _Atractylodes lancea_ (Thunb.) DC. rhizome exhibit synergistic effect against cholangiocarcinoma cell. _J. Exp. Pharmacol._ 10, 59–64

(2018). Article  PubMed  PubMed Central  CAS  Google Scholar  * Kotawong, K., Chaijaroenkul, W., Muhamad, P. & Na-Bangchang, K. Cytotoxic activities and effects of atractylodin and

β-eudesmol on the cell cycle arrest and apoptosis on cholangiocarcinoma cell line. _J. Pharmacol. Sci._ 136(2), 51–56. https://doi.org/10.1016/j.jphs.2017.09.033 (2018). Article  PubMed  CAS

  Google Scholar  * Srijiwangsa, P., Ponnikorn, S. & Na-Bangchang, K. Effect of β-eudesmol on NQO1 suppression-enhanced sensitivity of cholangiocarcinoma cells to chemotherapeutic

agents. _BMC Pharmacol. Toxicol._ 19(1), 32. https://doi.org/10.1186/s40360-018-0223-4 (2018). Article  PubMed  PubMed Central  CAS  Google Scholar  * Acharya, B., Chajaroenkul, W. &

Na-Bangchang, K. β-Eudesmol inhibits the migration of cholangiocarcinoma cells by suppressing epithelial-mesenchymal transition via PI3K/AKT and p38MAPK modulation. _Asian Pac. J. Cancer

Prev._ 23 (8), 2573–2581. https://doi.org/10.31557/APJCP.2022.23.8.2573 (2022). Article  PubMed  PubMed Central  CAS  Google Scholar  * Tshering, G., Pimtong, W., Plengsuriyakarn, T. &

Na-Bangchang, K. Anti-angiogenic effects of beta-eudesmol and atractylodin in developing zebrafish embryos. _Comp. Biochem. Physiol. C Toxicol. Pharmacol._ 243, 108980.

https://doi.org/10.1016/j.cbpc.2021.108980 (2021). Article  PubMed  CAS  Google Scholar  * Sonsomnuek, P., Tarasuk, M., Plengsuriyakarn, T., Boonprasert, K. & Na-Bangchang, K. Apoptotic

and anti-metastatic effects of _Atractylodes lancea_ (Thunb.) DC. in a Hamster model of cholangiocarcinoma. _Asian Pac. J. Cancer Prev._ 23 (9), 3093–3101.

https://doi.org/10.31557/APJCP.2022.23.9.3093 (2022). Article  PubMed  PubMed Central  CAS  Google Scholar  * Zuo, J. et al. Integrating molecular network and metabolomics study on

anti-rheumatic mechanisms and antagonistic effects against methotrexate-induced toxicity of Qing-Luo-Yin. _Front. Pharmacol._ 9, 1472. https://doi.org/10.3389/fphar.2018.01472 (2018). [old:

5]. Article  PubMed  PubMed Central  CAS  Google Scholar  * Liu, Z., Ma, H. & Lai, Z. Revealing the potential mechanism of _Astragalus membranaceus_ improving prognosis of hepatocellular

carcinoma by combining transcriptomics and molecular network. _BMC Complement. Med. Ther._ 21(1), 263. https://doi.org/10.1186/s12906-021-03425-9 (2021). Article  PubMed  PubMed Central 

CAS  Google Scholar  * Lu, Y., Dong, K., Yang, M. & Liu, J. Molecular network-based strategy to investigate the bioactive ingredients and molecular mechanism of _Evodia rutaecarpa_ in

colorectal cancer. _BMC Complement. Med. Ther._ 23, 1433. https://doi.org/10.1186/s12906-023-04254-8 (2023). Article  CAS  Google Scholar  * Na-Bangchang, K. et al. Phase IIa clinical trial

to evaluate safety and efficacy of capsule formulation of the standardized extract of _Atractylodes lancea_ (Thunb) DC in patients with advanced-stage intrahepatic cholangiocarcinoma. _J.

Tradit Complement. Med._ (2023). (In press). * Odongo, R., Demiroglu-Zergeroglu, A. & Çakır, T. A systems pharmacology approach based on oncogenic signalling pathways to determine the

mechanisms of action of natural products in breast cancer from transcriptome data. _BMC Complement. Med. Ther._ 21(1), 181. https://doi.org/10.1186/s12906-021-03340-z (2021). Article  PubMed

  PubMed Central  CAS  Google Scholar  * Zhou, Z. et al. Tumor-associated neutrophils and macrophages interaction contributes to intrahepatic cholangiocarcinoma progression by activating

STAT3. _J. Immunother Cancer_ 9(3), e001946. https://doi.org/10.1136/jitc-2020-001946 (2021). Article  PubMed  PubMed Central  Google Scholar  * Yothaisong, S. et al. Increased activation of

PI3K/AKT signaling pathway is associated with cholangiocarcinoma metastasis and PI3K/mTOR Inhibition presents a possible therapeutic strategy. _Tumour Biol._ 34(6), 3637–3648.

https://doi.org/10.1007/s13277-013-0945-2 (2013). Article  PubMed  CAS  Google Scholar  * Fang, Z., Lu, L., Tian, Z. & Luo, K. Overexpression of phosphorylated 4E-binding protein 1

predicts lymph node metastasis and poor prognosis of Chinese patients with hilar cholangiocarcinoma. _Med. Oncol._ 31, 5940. https://doi.org/10.1007/s12032-014-0940-5 (2014). Article  CAS 

Google Scholar  * Liao, W., Lin, J. X. & Leonard, W. J. Interleukin-2 at the crossroads of effector responses, tolerance, and immunotherapy. _Immunity_ 38(1), 13–25.

https://doi.org/10.1016/j.immuni.2013.01.004 (2013). Article  PubMed  PubMed Central  CAS  Google Scholar  * Biswas, P. et al. Analysis of SYK gene as a prognostic biomarker and suggested

potential bioactive phytochemicals as an alternative therapeutic option for colorectal cancer: an in-silico pharmaco-Informatics investigation. _J. Pers. Med._ 11 (9), 888.

https://doi.org/10.3390/jpm11090888 (2021). Article  PubMed  PubMed Central  Google Scholar  * Kang, Q. et al. Characterization and prognostic significance of mortalin, Bcl-2 and Bax in

intrahepatic cholangiocarcinoma. _Oncol. Lett._ 15, 2161–2168. https://doi.org/10.3892/ol.2017.7570 (2018). Article  PubMed  CAS  Google Scholar  * Chen, L. & Tiannqianng, S. Multi-omics

characterization of cholangiocarcinoma and association with prognostic and therapeutic molecular subtypes (abstract). _ASCO_ Supplemennt_1.60 (2023) * Xia, T. et al. Immune cell atlas of

cholangiocarcinomas reveals distinct tumor microenvironments and associated prognoses. _J. Hematol. Oncol._ 15(1), 37. https://doi.org/10.1186/s13045-022-01253-z (2022). Article  PubMed 

PubMed Central  CAS  Google Scholar  * Taniai, M. et al. Mcl-1 mediates tumor necrosis factor-related apoptosis-inducing ligand resistance in human cholangiocarcinoma cells. _Cancer Res._

64(10), 3517–3524. https://doi.org/10.1158/0008-5472.CAN-03-2770 (2004). Article  PubMed  CAS  Google Scholar  * Kobayashi, S., Werneburg, N. W., Bronk, S. F., Kaufmann, S. H. & Gores,

G. J. Interleukin-6 contributes to Mcl-1 up-regulation and TRAIL resistance via an Akt-signaling pathway in cholangiocarcinoma cells. _Gastroenterology_ 128(7), 2054–2065.

https://doi.org/10.1053/j.gastro.2005.03.010 (2005). Article  PubMed  CAS  Google Scholar  * Roy, S., Glaser, S. & Chakraborty, S. Inflammation and progression of cholangiocarcinoma:

Role of angiogenic and lymphangiogenic mechanisms. _Front. Med. (Lausanne)_ 6, 293. https://doi.org/10.3389/fmed.2019.00293 (2019). Article  PubMed  Google Scholar  * Ma, L. et al.

Single-cell atlas of tumor cell evolution in response to therapy in hepatocellular carcinoma and intrahepatic cholangiocarcinoma. _J. Hepatol._ 75, 1397–1408.

https://doi.org/10.1016/j.jhep.2021.06.028 (2021). Article  PubMed  PubMed Central  CAS  Google Scholar  * Vogler, M. BCL2A1: The underdog in the BCL2 family. _Cell. Death Differ._ 19,

67–74. https://doi.org/10.1038/cdd.2011.158 (2012). Article  PubMed  CAS  Google Scholar  * Pan, J. et al. Reactive oxygen species-activated akt/ask1/p38 signaling pathway in nickel

compound-induced apoptosis in Beas 2b cells. _Chem. Res. Toxicol._ 23, 568–557 (2010). Article  ADS  PubMed  PubMed Central  CAS  Google Scholar  * Donato, N. J. & Perez, M. Tumor

necrosis factor-induced apoptosis stimulates p53 accumulation and p21WAF1 proteolysis in ME-180 cells. _J. Biol. Chem._ 273(9), 5067–5072. https://doi.org/10.1074/jbc.273.9.5067 (1998).

Article  PubMed  CAS  Google Scholar  * Gámez-García, A. et al. ERK5 Inhibition induces autophagy-mediated cancer cell death by activating ER stress. _Front. Cell. Dev. Biol._ 9, 742049.

https://doi.org/10.3389/fcell.2021.742049 (2021). Article  PubMed  PubMed Central  Google Scholar  * Stoyanova, T., Roy, N., Kopanja, D., Bagchi, S. & Raychaudhuri, P. DDB2 decides cell

fate following DNA damage. _Proc. Natl. Acad. Sci. U S A_. 106(26), 10690–10695. https://doi.org/10.1073/pnas.0812254106 (2009). Article  ADS  PubMed  PubMed Central  Google Scholar  *

Prager, I. & Watzl, C. Mechanisms of natural killer cell-mediated cellular cytotoxicity. _J. Leukoc. Biol._ 105(6), 1319–1329. https://doi.org/10.1002/JLB.MR0718-269R (2019). Article 

PubMed  CAS  Google Scholar  * Moon, T. D., Morley, J. E., Vessella, R. L. & Lange, P. H. The role of calmodulin in human natural killer cell activity. _Scand. J. Immunol._ 18(3),

255–258. https://doi.org/10.1111/j.1365-3083.1983.tb00865.x (1983). Article  PubMed  CAS  Google Scholar  * Yan, Y., Gao, Z., Han, H., Zhao, Y. & Zhang, Y. NRAS expression is associated

with prognosis and tumor immune microenvironment in lung adenocarcinoma. _J. Cancer Res. Clin. Oncol._ 148(3), 565–575. https://doi.org/10.1007/s00432-021-03842- (2022). Article  PubMed  CAS

  Google Scholar  * Poltorak, M., Meinert, I., Stone, J. C., Schraven, B. & Simeoni, L. Sos1 regulates sustained TCR-mediated Erk activation. _Eur. J. Immunol._ 44(5), 1535–1540.

https://doi.org/10.1002/eji.201344046 (2014). Article  PubMed  CAS  Google Scholar  * Vaquero, J., Lobe, C. & Fouassier, L. Unveiling resistance mechanisms to EGFR inhibitors in

cholangiocarcinoma. _Oncotarget_ 9, 37274–37275. https://doi.org/10.18632/oncotarget.26403 (2018). Article  PubMed  PubMed Central  Google Scholar  * Peraldo-Neia, C. et al. Prognostic and

predictive role of EGFR pathway alterations in biliary cancer patients treated with chemotherapy and anti-EGFR. _PLoS One_ 13(1), e0191593. https://doi.org/10.1371/journal.pone.0191593

(2018). Article  PubMed  PubMed Central  CAS  Google Scholar  * Yoon, H., Min, J. K., Lee, J. W., Kim, D. G. & Hong, H. J. Acquisition of chemoresistance in intrahepatic

cholangiocarcinoma cells by activation of AKT and extracellular signal-regulated kinase (ERK)1/2. _Biochem. Biophys. Res. Commun._ 405, 3. https://doi.org/10.1016/j.bbrc.2010.11.130 (2011).

Article  CAS  Google Scholar  * De, S., Dermawan, J. K. & Stark, G. R. EGF receptor uses SOS1 to drive constitutive activation of NFκB in cancer cells. _Proc. Natl. Acad. Sci. U S A_.

111(32), 11721–11726. https://doi.org/10.1073/pnas.1412390111 (2014). Article  ADS  PubMed  PubMed Central  CAS  Google Scholar  * Loilome, W. et al. PRKAR1A overexpression is associated

with increased ECPKA autoantibody in liver fluke-associated cholangiocarcinoma: application for assessment of the risk group. _Tumour Biol._ 33(6), 2289–2298.

https://doi.org/10.1007/s13277-012-0491-3 (2012). Article  PubMed  CAS  Google Scholar  * Zheng, Y. et al. Specific genomic alterations and prognostic analysis of perihilar

cholangiocarcinoma and distal cholangiocarcinoma. _J. Gastrointest. Oncol._ 12(6), 2631–2642. https://doi.org/10.21037/jgo-21-776 (2021). Article  PubMed  PubMed Central  Google Scholar  *

Ma, D., Lian, F. & Wang, X. PLCG2 promotes hepatocyte proliferation in vitro via NF-κB and ERK pathway by targeting bcl2, Myc and ccnd1. _Artif. Cells Nanomed. Biotechnol._ 47,

1:3786–3792. https://doi.org/10.1080/21691401.2019.1669616 (2019). Article  PubMed  CAS  Google Scholar  * Alvaro, D. et al. Estrogens and insulin-like growth factor 1 modulate neoplastic

cell growth in human cholangiocarcinoma. _Am. J. Pathol._ 169(3), 877–888. https://doi.org/10.2353/ajpath.2006.050464 (2006). Article  PubMed  PubMed Central  CAS  Google Scholar  * Vaquero,

J. et al. The IGF2/IR/IGF1R pathway in tumor cells and myofibroblasts mediates resistance to EGFR Inhibition in cholangiocarcinoma. _Clin. Cancer Res._ 24(17), 4282–4296.

https://doi.org/10.1158/1078-0432.CCR-17-3725 (2018). Article  PubMed  CAS  Google Scholar  * Matallanas, D. et al. Raf family kinases: old dogs have learned new tricks. _Genes Cancer_ 2(3),

232–260. https://doi.org/10.1177/1947601911407323 (2011). Article  PubMed  PubMed Central  CAS  Google Scholar  * West, N. R., Owens, B. M. J. & Hegazy, A. N. The oncostatin M-stromal

cell axis in health and disease. _Scand. J. Immunol._ 88(3), e12694. https://doi.org/10.1111/sji.12694 (2018). Article  PubMed  CAS  Google Scholar  * Dai, J. et al. SPOP regulates the

expression profiles and alternative splicing events in human hepatocytes. _Open. Life Sci._ 18(1), 20220755. https://doi.org/10.1515/biol-2022-0755 (2023). Article  PubMed  PubMed Central 

CAS  Google Scholar  * Ali, A., Nimisha, Sharma, A. K., Mishra, P. K. & Saluja, S. S. Clinical significance of SPOP and APC gene alterations in colorectal cancer in Indian population.

_Mol. Genet. Genomics_ 298(5), 1087–1105. https://doi.org/10.1007/s00438-023-02029-x (2023). Article  PubMed  CAS  Google Scholar  * Xu, J. et al. Properties and clinical relevance of

speckle-type POZ protein in human colorectal cancer. _J. Gastrointest. Surg._ 19, 1484–1496. https://doi.org/10.1007/s11605-015-2767-6 (2015). Article  PubMed  Google Scholar  * Zhao, W.,

Zhou, J., Deng, Z., Gao, Y. & Cheng, Y. SPOP promotes tumor progression via activation of beta-catenin/TCF4 complex in clear cell renal cell carcinoma. _Int. J. Oncol._ 49, 1001–1008.

https://doi.org/10.3892/ijo.2016.3609 (2016). Article  PubMed  CAS  Google Scholar  * Wan, D. et al. PRKAR2A-derived circular RNAs promote the malignant transformation of colitis and

distinguish patients with colitis-associated colorectal cancer. _Clin. Transl Med._ 12 (2), e683. https://doi.org/10.1002/ctm2.683 (2022). Article  PubMed  PubMed Central  CAS  Google

Scholar  * Duan, T., Zhou, D., Yao, Y. & Shao, X. The association of aberrant expression of FGF1 and mTOR-S6K1 in colorectal Cancer. _Front. Oncol._ 11, 706838.

https://doi.org/10.3389/fonc.2021.706838 (2021). Erratum in: Front Oncol. 2021;11:792453. Article  PubMed  PubMed Central  CAS  Google Scholar  * Mehta, A. K., Gracias, D. T. & Croft, M.

TNF activity and T cells. _Cytokine_ 101, 14–18. https://doi.org/10.1016/j.cyto.2016.08.003 (2018). * Uramoto, H., Akyürek, L. M. & Hanagiri, T. A positive relationship between filamin

and VEGF in patients with lung cancer. _Anticancer Res._ 30, 10:3939–3944 (2010). PubMed  Google Scholar  * Huang, Y. et al. High expression of ITGA3 promotes proliferation and cell cycle

progression and indicates poor prognosis in intrahepatic cholangiocarcinoma. _Biomed. Res. Int._ 2018, 2352139. https://doi.org/10.1155/2018/2352139 (2018). Article  PubMed  PubMed Central 

CAS  Google Scholar  * Ning, P. et al. PLAU plays a functional role in driving lung squamous cell carcinoma metastasis. _Genes Dis._ 11(2), 554–557.

https://doi.org/10.1016/j.gendis.2023.04.010 (2023). Article  PubMed  PubMed Central  CAS  Google Scholar  * Wen, S. C. et al. ets-2 regulates cdc2 kinase activity in mammalian cells:

coordinated expression of cdc2 and Cyclin A. _Exp. Cell. Res._ 217(1), 8–14. https://doi.org/10.1006/excr.1995.1057 (1995). Article  PubMed  CAS  Google Scholar  * Zabuawala, T. et al. An

ets2-driven transcriptional program in tumor-associated macrophages promotes tumor metastasis. _Cancer Res._ 70(4), 1323–1333. https://doi.org/10.1158/0008-5472.CAN-09-1474 (2010). Article 

PubMed  PubMed Central  CAS  Google Scholar  * Szymczyk, J. et al. FGF1 protects FGFR1-overexpressing cancer cells against drugs targeting tubulin polymerization by activating AKT via two

independent mechanisms. _Front. Oncol._ 12, 1011762. https://doi.org/10.3389/fonc.2022.1011762 (2022). Article  PubMed  PubMed Central  CAS  Google Scholar  * Hong, J. et al. Expression of

variant isoforms of the tyrosine kinase SYK determines the prognosis of hepatocellular carcinoma. _Cancer Res._ 74, 6:1845–1856. https://doi.org/10.1158/0008-5472.CAN-13-2104 (2014). Article

  PubMed  PubMed Central  CAS  Google Scholar  * Oba, M. et al. CCR7 mediates cell invasion and migration in extrahepatic cholangiocarcinoma by inducing epithelial-mesenchymal transition.

_Cancers (Basel)_ 15, 61878. https://doi.org/10.3390/cancers15061878 (2023). Article  CAS  Google Scholar  * Roy, S., Glaser, S. & Chakraborty, S. Inflammation and progression of

cholangiocarcinoma: role of angiogenic and lymphangiogenic mechanisms. _Front. Med. (Lausanne)_ 6, 293. https://doi.org/10.3389/fmed.2019.00293 * Taniai, M. et al. Mcl-1 mediates tumor

necrosis factor-related apoptosis-inducing ligand resistance in human cholangiocarcinoma cells. _Cancer Res._ 64, 10 (2004). Article  Google Scholar  * Ma, B. et al. Distinct clinical and

prognostic implication of IDH1/2 mutation and other most frequent mutations in large duct and small duct subtypes of intrahepatic cholangiocarcinoma. _BMC Cancer_ 20(1), 318.

https://doi.org/10.1186/s12885-020-06804-6 (2020). Article  PubMed  PubMed Central  CAS  Google Scholar  * Mehta, A. K., Gracias, D. T. & Croft, M. TNF activity and T cells. _Cytokine_

101, 14–18. https://doi.org/10.1016/j.cyto.2016.08.003 (2018). Article  PubMed  CAS  Google Scholar  * Ran., G. H. et al. Natural killer cell homing and trafficking in tissues and tumors:

From biology to application. _Signal. Transduct. Target. Ther._ 7(1), 205. https://doi.org/10.1038/s41392-022-01058-z (2022). Article  PubMed  PubMed Central  CAS  Google Scholar  * Wang,

J., Jiang, Y. H., Yang, P. Y., Liu, F. & Increased Collagen Type, V. α2 (COL5A2) in colorectal Cancer is associated with poor prognosis and tumor progression. _Onco Targets Ther._ 14,

2991–3002. https://doi.org/10.2147/OTT.S288422 (2021). Article  PubMed  PubMed Central  Google Scholar  * Liu, Q. et al. Oncostatin M expression and TP53 mutation status regulate

tumor-infiltration of immune cells and survival outcomes in cholangiocarcinoma. _Aging (Albany NY)_ 12(21), 21518–21543. https://doi.org/10.18632/aging.103936 (2020). Article  PubMed  CAS 

Google Scholar  * Chen, W. et al. Unraveling the heterogeneity of cholangiocarcinoma and identifying biomarkers and therapeutic strategies with single-cell sequencing technology. _Biomed.

Pharmacother._ 162, 114697. https://doi.org/10.1016/j.biopha.2023.114697 (2023). Article  PubMed  CAS  Google Scholar  * Lin, P. et al. Survival analysis of genome-wide profiles coupled with

connectivity map database mining to identify potential therapeutic targets for cholangiocarcinoma. _Oncol. Rep._ 40(6), 3189–3198. https://doi.org/10.3892/or.2018.6710 (2018). Article 

PubMed  PubMed Central  CAS  Google Scholar  * Abell, K. & Watson, C. J. The Jak/Stat pathway: A novel way to regulate PI3K activity. _Cell. Cycle_ 4, 897–900.

https://doi.org/10.4161/cc.4.7.1837 (2005). Article  PubMed  CAS  Google Scholar  * Tang, S. et al. Association analyses of the JAK/STAT signaling pathway with the progression and prognosis

of colon cancer. _Oncol. Lett._ 17(1), 159–164. https://doi.org/10.3892/ol.2018.9569 (2019). Article  PubMed  CAS  Google Scholar  * Chen, B. et al. JAK1 as a prognostic marker and its

correlation with immune infiltrates in breast cancer. _Aging (Albany NY)_. 11(23), 11124–11135. https://doi.org/10.18632/aging.102514 (2019). Article  PubMed  CAS  Google Scholar  * Pei, Y.,

Cui, X. & Wang, Y. Regulation of IL-10 expression and function by JAK-STAT in CD8 + T cells. _Int. Immunopharmacol._ 128, 111563. https://doi.org/10.1016/j.intimp.2024.111563 (2024).

Article  PubMed  CAS  Google Scholar  * Shen, K. et al. The expression landscape of JAK1 and its potential as a biomarker for prognosis and immune infiltrates in NSCLC. _BMC Bioinform._ 22,

1471. https://doi.org/10.1186/s12859-021-04379-y (2021). Article  CAS  Google Scholar  * Suksawat, M. et al. Upregulation of endothelial nitric oxide synthase (eNOS) and its upstream

regulators in _Opisthorchis viverrini_ associated cholangiocarcinoma and its clinical significance. _Parasitol. Int._ 66, 4:486–493. https://doi.org/10.1016/j.parint.2016.04.008 (2017).

Article  PubMed  CAS  Google Scholar  * Alwithenani, A. I. & Althubiti, M. A. Systematic analysis of spleen tyrosine kinase expression and its clinical outcomes in various cancers.

_Saudi J. Med. Med. Sci._ 8(2), 95–104. https://doi.org/10.4103/sjmms.sjmms_300_19 (2020). Article  PubMed  PubMed Central  Google Scholar  * Guittard, G. et al. Absence of both Sos-1 and

Sos-2 in peripheral CD4(+) T cells leads to PI3K pathway activation and defects in migration. _Eur. J. Immunol._ 45, 8. https://doi.org/10.1002/eji.201445226 (2015). Article  CAS  Google

Scholar  * Alem, D. et al. Translational relevance of SOS1 targeting for KRAS-mutant colorectal cancer. _Mol. Carcinog._ 62(7), 1025–1037. https://doi.org/10.1002/mc.23543 (2023). Article 

PubMed  PubMed Central  CAS  Google Scholar  * Li, Y., Yin, Y., He, Y., He, K. & Li, J. SOS1 regulates HCC cell epithelial-mesenchymal transition via the PI3K/AKT/mTOR pathway. _Biochem.

Biophys. Res. Commun._ 637, 161–169. https://doi.org/10.1016/j.bbrc.2022.11.015 (2022). Article  PubMed  CAS  Google Scholar  * Li, Q. et al. High integrin Α3 expression is associated with

poor prognosis in patients with non-small cell lung cancer. _Transl Lung Cancer Res._ 9(4), 1361–1378. https://doi.org/10.21037/tlcr-19-633 (2020). Article  PubMed  PubMed Central  CAS 

Google Scholar  * Wu, A. et al. Integrated analysis of prognostic and immune associated integrin family in ovarian cancer. _Front. Genet._ 11, 705. https://doi.org/10.3389/fgene.2020.00705

(2020). Article  PubMed  PubMed Central  CAS  Google Scholar  * Zheng, Q., Zhang, B., Li, C. & Zhang, X. Overcome drug resistance in cholangiocarcinoma: New insight into mechanisms and

refining the preclinical experiment models. _Front. Oncol._ 12, 850732. https://doi.org/10.3389/fonc.2022.850732 (2022). Article  PubMed  PubMed Central  CAS  Google Scholar  * Sridharan, V.

et al. FGFR mRNA expression in cholangiocarcinoma and its correlation with FGFR2 fusion status and immune signatures. _Clin. Cancer Res._ 28(24), 5431–5439.

https://doi.org/10.1158/1078-0432.CCR-22-1244 (2022). Article  PubMed  PubMed Central  CAS  Google Scholar  * Krook, M. A. et al. Efficacy of FGFR inhibitors and combination therapies for

acquired resistance in FGFR2-fusion cholangiocarcinoma. _Mol. Cancer Ther._ 19(3), 847–857. https://doi.org/10.1158/1535-7163.MCT-19-0631 (2020). Article  PubMed  PubMed Central  CAS  Google

Scholar  * Zamai, L. et al. Natural killer (NK) cell-mediated cytotoxicity: differential use of TRAIL and fas ligand by immature and mature primary human NK cells. _J. Exp. Med._ 188(12),

2375–2380. https://doi.org/10.1084/jem.188.12.2375 (1998). Article  PubMed  PubMed Central  CAS  Google Scholar  * Matsuzawa, A. et al. ROS-dependent activation of the TRAF6-ASK1-p38 pathway

is selectively required for TLR4-mediated innate immunity. _Nat. Immunol._ 6(6), 587–592. https://doi.org/10.1038/ni1200 (2005). Article  PubMed  CAS  Google Scholar  * Ghafouri-Fard, S. et

al. 5-Fluorouracil: A narrative review on the role of regulatory mechanisms in driving resistance to this chemotherapeutic agent. _Front. Oncol._ 11, 658636 (2021). Article  PubMed  PubMed

Central  CAS  Google Scholar  * Taïeb, J. et al. Optimization of 5-fluorouracil (5-FU)/cisplatin combination chemotherapy with a new schedule of leucovorin, 5-FU and cisplatin (LV5FU2-P

regimen) in patients with biliary tract carcinoma. _Ann. Oncol._ 13(8), 1192–1196 (2002). Article  PubMed  Google Scholar  * Siebenhüner, A. R. et al. Adjuvant treatment of resectable

biliary tract cancer with cisplatin plus gemcitabine: A prospective single center phase II study. _BMC Cancer_18(1), 72 (2018). Article  PubMed  PubMed Central  Google Scholar  * Kotawong,

K., Chajaroenkul, W., Roytrakul, S., Phaonakrop, N. & Na-Bangchang, K. The proteomics and metabolomics analysis for screening the molecular targets of action of beta-eudesmol in

cholangiocarcinoma. _Asian Pac. J. Cancer Prev._ 22, 3:909–918. https://doi.org/10.31557/APJCP.2021.22.3.909 (2021). Article  PubMed  PubMed Central  CAS  Google Scholar  * Mathema, V. B.,

Chaijaroenkul, W., Karbwang, J. & Na-Bangchang, K. Growth inhibitory effect of β-eudesmol on cholangiocarcinoma cells and its potential suppressive effect on heme oxygenase-1 production,

STAT1/3 activation, and NF-κB downregulation. _Clin. Exp. Pharmacol. Physiol._ 44(11), 1145–1154. https://doi.org/10.1111/1440-1681.12818 (2017). Article  PubMed  CAS  Google Scholar 

Download references FUNDING This research project was supported by the Thailand Science Research and Innovation Fundamental Fund Fiscal Year 2023 and Thammasat University (Chulabhorn

International College of Medicine, Center of Excellence in Pharmacology and Molecular Biology of Malaria and Cholangiocarcinoma). Kesara Na-Bangchang is funded by the National Research

Council of Thailand (NRCT): Contract number N42A671041. All funders have no roles for publication. AUTHOR INFORMATION AUTHORS AND AFFILIATIONS * Centre of Excellence in Pharmacology and

Molecular Biology of Malaria and Cholangiocarcinoma, Chulabhorn International College of Medicine, Thammasat University (Rangsit Campus), 99 moo 18, Phaholyothin Road, Klong Luang District,

Pathumthani, 12121, Thailand Teerachat Saeheng, Wanna Chaijaorenkul & Kesara Na-Bangchang * Gradulate Program in Bioclinical Science, Chulabhorn International College of Medicine,

Thammasat University (Rangsit Campus), Klong Luang District, Pathumtanee province, Thailand Teerachat Saeheng, Ethan Vindvamara, Wanna Chaijaorenkul & Kesara Na-Bangchang * Sakol Nakorn

Hospital, Sakol Nakorn, Sakol Nakorn Province, Thailand Nisit Tongsiri & Kesara Na-Bangchang Authors * Teerachat Saeheng View author publications You can also search for this author

inPubMed Google Scholar * Ethan Vindvamara View author publications You can also search for this author inPubMed Google Scholar * Wanna Chaijaorenkul View author publications You can also

search for this author inPubMed Google Scholar * Nisit Tongsiri View author publications You can also search for this author inPubMed Google Scholar * Kesara Na-Bangchang View author

publications You can also search for this author inPubMed Google Scholar CONTRIBUTIONS T.S.: Conceptualization, data curation, validation, formal analysis, methodology, writing-original

draft, interpretation. E.V.: Experimental study, W.C.: Experimental study. N.T.: Experimental study. K.N.: Conceptualization, data curation, funding acquisition, supervision, project

administration, writing—review, and editing. All authors reviewed the manuscript. CORRESPONDING AUTHOR Correspondence to Kesara Na-Bangchang. ETHICS DECLARATIONS COMPETING INTERESTS The

authors declare no competing interests. ETHICAL APPROVAL AND CONSENT TO PARTICIPATE This project has approval from the Human Research Ethics Committee of Sakol Nakorn hospital (No.

049/2563). All participants signed written informed consent forms prior to enrollment. ADDITIONAL INFORMATION PUBLISHER’S NOTE Springer Nature remains neutral with regard to jurisdictional

claims in published maps and institutional affiliations. ELECTRONIC SUPPLEMENTARY MATERIAL Below is the link to the electronic supplementary material. SUPPLEMENTARY MATERIAL 1. RIGHTS AND

PERMISSIONS OPEN ACCESS This article is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License, which permits any non-commercial use, sharing,

distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and

indicate if you modified the licensed material. You do not have permission under this licence to share adapted material derived from this article or parts of it. The images or other third

party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the

article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright

holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by-nc-nd/4.0/. Reprints and permissions ABOUT THIS ARTICLE CITE THIS ARTICLE Saeheng, T., Vindvamara, E.,

Chaijaorenkul, W. _et al._ A molecular network analysis and in silico docking of beta-eudesmol, atractylodin and hinesol in patients with advance stage intrahepatic cholangiocarcinoma. _Sci

Rep_ 15, 16279 (2025). https://doi.org/10.1038/s41598-025-91968-z Download citation * Received: 10 May 2024 * Accepted: 24 February 2025 * Published: 10 May 2025 * DOI:

https://doi.org/10.1038/s41598-025-91968-z SHARE THIS ARTICLE Anyone you share the following link with will be able to read this content: Get shareable link Sorry, a shareable link is not

currently available for this article. Copy to clipboard Provided by the Springer Nature SharedIt content-sharing initiative KEYWORDS * Molecular network * _Atractylodes lancea_ Thunb (DC.) *

Traditional medicine * Cholangiocarcinoma