Essentiality-driven prediction of anticancer drug responses in preclinical and clinical contexts

Explainable Multi-Omic Machine Learning Framework for Predicting Drug Response in Breast Cancer

Accurate prediction of drug sensitivity in cancer cell lines is vital for precision oncology and patient-specific therapies. However, many computational approaches fail to integrate multi-modal biological and chemical features and often struggle with high-dimensional, imbalanced pharmacogenomic data, limiting predictive accuracy and interpretability. To address these challenges, we developed a machine learning framework that integrates pharmacogenomic profiles-including mutation status, copy number alterations, and microsatellite instabil-ity-with molecular fingerprints and descriptors of 85 anticancer drugs, generated using PaDEL from SMILES strings. Data from 40 breast cancer cell lines in the Genomics of Drug Sensitivity in Cancer (GDSC) dataset were employed. A threestage feature selection strategy combining Boruta, mRMR, and XGBoost was applied to reduce drug feature dimensionality while retaining 130 cell line features. Multiple models were trained, and LightGBM, optimized with grid search, class weighting, and 3-fold cross-validation, demonstrated superior performance in handling severe class imbalance (233 sensitive vs. 3167 resistant samples). LightGBM achieved training AUROC $=0.9455$, AUPRC $\boldsymbol{=} \mathbf{0. 5 1 4 8}$, Accuracy $\boldsymbol{=} \mathbf{0. 8 4 1 5}$, F1-score = 0.4481, Recall = 0.9409, and MCC = 0.4732, underscoring its suitability for sparse biomedical datasets. Model interpretation with SHapley Additive exPlanations (SHAP) highlighted BRCA-related features, identifying cnaBRCA25 (not mutated) as a resistance marker and cnaBRCA47 (mutated) as a context-dependent biomarker, consistent with their roles in DNA repair pathways. Overall, this framework demonstrates the value of multi-modal integration and interpretable machine learning in pharmacogenomics. While results are promising, validation on larger and independent cohorts is essential to establish clinical relevance.

Epigenetic profile drives accurate survival prediction in breast cancer via a multi-omics machine learning model

Bioinformatics-based identification of glycolysis-related signatures associated with drug resistance and prognosis in lung adenocarcinoma

Background Drug resistance and poor clinical outcomes in lung adenocarcinoma (LUAD) necessitate robust biomarkers for personalized therapy. Glycolysis reprogramming is a hallmark of cancer, but its clinical utility remains incompletely defined. Methods We integrated TCGA and GEO transcriptomic data with Weighted gene co−expression network analysis (WGCNA), least absolute shrinkage and selection operator (LASSO), and multivariate Cox regression to construct a glycolysis−related prognostic signature. A nomogram combining the risk score with clinicopathological factors was developed. Drug sensitivity was predicted using the pRRophetic algorithm. qRT−PCR and xenograft models using A549 and cisplatin−resistant A549/DDP cells validated the expression of candidate genes. Results Patients stratified by glycolysis-related risk scores exhibited significantly distinct survival outcomes, and the glycolysis-based signature functioned as an independent prognostic factor for overall survival in LUAD. The nomogram demonstrated robust predictive performance and effectively estimated patient sensitivity to three commonly used conventional chemotherapeutic agents. In vitro and in vivo studies using A549 cells and their cisplatin-resistant derivative A549/DDP revealed aberrant expression of VIPR1, ADRB2, RXFP1, PDGFB, WNT3A, and SPRY1 in the resistant cell line and in corresponding xenograft tumor tissues. These findings suggest that glycolytic activity is closely associated with both drug resistance and clinical prognosis in LUAD. Conclusions This study identifies a glycolysis-related gene signature with demonstrable utility for prognostic stratification and therapeutic response prediction in LUAD. The proposed integrative model holds promise for enhancing precision treatment decision-making through optimized risk assessment and rational selection of chemotherapeutic regimens.

NELLY enables patient-centric drug prioritization through interpretable drug-conditioned gene weighting

Precision oncology seeks to match each tumor with the most effective anti-cancer therapy. Advances in pharmacogenomics and machine learning enabled drug response prediction models with strong performance in cancer cell lines. Nonetheless, patient-centric evaluation of drug prioritization and systematic assessment of model generalization in patient-derived systems across cancer types remain largely absent. Here we introduce a translational framework combining patient-centric benchmarking with a pan-cancer pharmacogenomic atlas of patient-derived organoids, together with NELLY, a deep learning model integrating transcriptomic and chemical information to predict drug response and prioritize therapies. NELLY outperformed existing methods for patient-specific drug prioritization across cancer cell lines and patient-derived organoids, including under out-of-distribution evaluation. Its dynamic weighting mechanism provided patient-specific gene attributions, offering a route to connect predicted drug response to molecular programs associated with drug resistance. Our results support NELLY as a promising framework for translationally relevant and interpretable drug response prediction in precision oncology.

Deep Learning-Driven Anticancer Drug Discovery: Emodepside as a Potential Therapeutic Candidate for Triple-Negative Breast Cancer

Triple-negative breast cancer (TNBC) is an aggressive breast cancer subtype with a poor prognosis. The absence of effective targeted therapies and endocrine treatment options leads to limited therapeutic options, which remains one of the major clinical challenges in TNBC management. Drug discovery is typically a lengthy and costly process that could be significantly improved through drug repurposing. However, the biological complexity and insufficient repurposing strategies hinder the reuse. This study aims to develop a deep learning-based framework to accelerate drug discovery for TNBC, identify novel therapeutic candidates, and uncover potential drug targets. We developed a deep neural network framework to predict the anticancer efficacy, toxicity profiles, and structural similarities of compounds. By applying this platform to screen over 6,000 compounds from the Drug Repurposing Hub, we identified promising candidates with potential therapeutic efficacy and safety profiles against TNBC. The top-predicted compounds were subsequently validated through comprehensive in vitro and in vivo functional assays. Furthermore, we employed transcriptomic sequencing and mass spectrometry-based proteomics to elucidate the molecular mechanisms underlying the anti-TNBC activity. We identified emodepside, a structurally unique molecule diverging from conventional anticancer agents that exhibited potent antitumor efficacy across multiple TNBC cell lines. Significantly, emodepside administration (5 mg/kg) inhibited tumor growth in xenograft models. Integrated multiomics analyses (RNA-seq/CETSA-MS) identified NAMPT as the primary target. This study demonstrates the viability of our deep learning models to discover structurally novel anticancer agents that are distinct from conventional drugs, thereby expanding the therapeutic arsenal for TNBC patients. Emodepside emerges as a promising TNBC therapeutic candidate, with a possible mechanism of promoting TNBC cell apoptosis via NAMPT inhibition.

Triptolide targets JUN to reverse cisplatin resistance of ovarian cancer: insights from single-cell transcriptome analysis and machine learning validation

Background Platinum-resistant ovarian cancer (PROC) is a major clinical challenge driven by profound intratumoral heterogeneity. Triptolide (TP) exhibits promising anti-tumor potential, yet its precise mechanisms within PROC remain elusive due to the limitations of traditional target-screening strategies. Methods This study developed a comprehensive strategy integrating computational predictions with in vitro experimental validations. First, scRNA-seq data were processed to evaluate TP target genes predicted by SwissTargetPrediction, alongside pseudotime trajectory and CellChat intercellular communication analyses. Subsequently, an ensemble of six machine learning (ML) algorithms (LASSO, Random Forest, Boruta, Decision Tree, XGBoost, and GBM) was utilized to pinpoint the core therapeutic target. To verify direct molecular engagement, molecular docking, molecular dynamics (MD) simulations, and Surface Plasmon Resonance (SPR) assays were performed. Finally, the functional mechanism of the identified target in CDDP resistance was validated in vitro using parental SKOV3 and resistant SKOV3/CDDP cell lines. Results scRNA-seq analysis revealed TP target genes are preferentially enriched in highly genomically unstable malignant epithelial cells. This subpopulation showed an aggressive intercellular communication profile, profound dependence on extracellular matrix (ECM) signals, and dominant secretion of the chemoresistance-related cytokine osteopontin (SPP1). Furthermore, the ML pipeline consistently pinpointed the proto-oncogene JUN as the core therapeutic target. Experiments confirmed TP efficiently suppressed aberrant c-Jun overexpression. Targeted JUN knockdown restored CDDP sensitivity, while its overexpression antagonized the synergistic cytotoxic and apoptotic effects of TP and CDDP. Ultimately, TP reverses CDDP resistance in PROC by downregulating JUN, dismantling intracellular pro-survival networks, and disrupting pro-tumorigenic crosstalk. Conclusion In conclusion, TP reverses CDDP resistance in PROC by downregulating JUN, which dismantles intracellular pro-survival networks. Integrating scRNA-seq and ML provides an accurate paradigm for deciphering botanical pharmacology, laying a strong foundation for the future development of TP-based therapies tailored for PROC.