MCMVDRP: a multi-channel multi-view deep learning framework for cancer drug response prediction.

Drug therapy remains the primary approach to treating tumours. Variability among cancer patients, including variations in genomic profiles, often results in divergent therapeutic responses to analogous anti-cancer drug treatments within the same cohort of cancer patients. Hence, predicting the drug response by analysing the genomic profile characteristics of individual patients holds significant research importance. With the notable progress in machine learning and deep learning, many effective methods have emerged for predicting drug responses utilizing features from both drugs and cell lines. However, these methods are inadequate in capturing a sufficient number of features inherent to drugs. Consequently, we propose a representational approach for drugs that incorporates three distinct types of features: the molecular graph, the SMILE strings, and the molecular fingerprints. In this study, a novel deep learning model, named MCMVDRP, is introduced for the prediction of cancer drug responses. In our proposed model, an amalgamation of these extracted features is performed, followed by the utilization of fully connected layers to predict the drug response based on the IC50 values. Experimental results demonstrate that the presented model outperforms current state-of-the-art models in performance.

Excavating regulated cell death signatures to predict prognosis, tumor microenvironment and therapeutic response in HR+/HER2- breast cancer.

Regulated cell death (RCD) has been documented to have great potentials for discovering novel biomarkers and therapeutic targets in malignancies. But its role and clinical value in HR+/HER2- breast cancer, the most common subtype of breast cancer, are obscure. In this study, we comprehensively explored 12 types of RCD patterns and found extensive mutations and dysregulations of RCD genes in HR+/HER2- breast cancer. A prognostic RCD scoring system (CDScore) based on six critical genes (LEF1, SLC7A11, SFRP1, IGFBP6, CXCL2, STXBP1) was constructed, in which a high CDScore predicts poor prognosis. The expressions and prognostic value of LEF1 and SFRP1were also validated in our tissue microarrays. The nomogram established basing on CDScore, age and TNM stage performed satisfactory in predicting overall survival, with an area under the ROC curve of 0.89, 0.82 and 0.8 in predicting 1-year, 3-year and 5-year overall survival rates, respectively. Furthermore, CDScore was identified to be correlated with tumor microenvironments and immune checkpoints by excavation of bulk and single-cell sequencing data. Patients in CDScore high group might be resistant to standard chemotherapy and target therapy. Our results underlined the potential effects and importance of RCD in HR+/HER2- breast cancer and provided novel biomarkers and therapeutic targets for HR+/HER2- breast cancer patients.

Methods in cancer research: Assessing therapy response of spheroid cultures by life cell imaging using a cost-effective live-dead staining protocol.

Spheroid cultures of cancer cell lines or primary cells represent a more clinically relevant model for predicting therapy response compared to two-dimensional cell culture. However, current live-dead staining protocols used for treatment response in spheroid cultures are often expensive, toxic to the cells, or limited in their ability to monitor therapy response over an extended period due to reduced stability. In our study, we have developed a cost-effective method utilizing calcein-AM and Helix NP™ Blue for live-dead staining, enabling the monitoring of therapy response of spheroid cultures for up to 10 days. Additionally, we used ICY BioImage Analysis and Z-stacks projection to calculate viability, which is a more accurate method for assessing treatment response compared to traditional methods on spheroid size. Using the example of glioblastoma cell lines and primary glioblastoma cells, we show that spheroid cultures typically exhibit a green outer layer of viable cells, a turquoise mantle of hypoxic quiescent cells, and a blue core of necrotic cells when visualized using confocal microscopy. Upon treatment of spheroids with the alkylating agent temozolomide, we observed a reduction in the viability of glioblastoma cells after an incubation period of 7 days. This method can also be adapted for monitoring therapy response in different cancer systems, offering a versatile and cost-effective approach for assessing therapy efficacy in three-dimensional culture models.

Characterization of Expression-Based Gene Clusters Gives Insights into Variation in Patient Response to Cancer Therapies.

Background: Transcriptomics can reveal much about cellular activity, and cancer transcriptomics have been useful in investigating tumor cell behaviors. Patterns in transcriptome-wide gene expression can be used to investigate biological mechanisms and pathways that can explain the variability in patient response to cancer therapies. Methods: We identified gene expression patterns related to patient drug response by clustering tumor gene expression data and selecting from the resulting gene clusters those where expression of cluster genes was related to patient survival on specific drugs. We then investigated these gene clusters for biological meaning using several approaches, including identifying common genomic locations and transcription factors whose targets were enriched in these clusters and performing survival analyses to support these candidate transcription factor-drug relationships. Results: We identified gene clusters related to drug-specific survival, and through these, we were able to associate observed variations in patient drug response to specific known biological phenomena. Specifically, our analysis implicated 2 stem cell-related transcription factors, HOXB4 and SALL4, in poor response to temozolomide in brain cancers. In addition, expression of SNRNP70 and its targets were implicated in cetuximab response by 3 different analyses, although the mechanism remains unclear. We also found evidence that 2 cancer-related chromosomal structural changes may impact drug efficacy. Conclusion: In this study, we present the gene clusters identified and the results of our systematic analysis linking drug efficacy to specific transcription factors, which are rich sources of potential mechanistic relationships impacting patient outcomes. We also highlight the most promising of these results, which were supported by multiple analyses and by previous research. We report these findings as promising avenues for independent validation and further research into cancer treatments and patient response.

Kinome state is predictive of cell viability in pancreatic cancer tumor and cancer-associated fibroblast cell lines.

Numerous aspects of cellular signaling are regulated by the kinome-the network of over 500 protein kinases that guides and modulates information transfer throughout the cell. The key role played by both individual kinases and assemblies of kinases organized into functional subnetworks leads to kinome dysregulation driving many diseases, particularly cancer. In the case of pancreatic ductal adenocarcinoma (PDAC), a variety of kinases and associated signaling pathways have been identified for their key role in the establishment of disease as well as its progression. However, the identification of additional relevant therapeutic targets has been slow and is further confounded by interactions between the tumor and the surrounding tumor microenvironment. In this work, we attempt to link the state of the human kinome, or kinotype, with cell viability in treated, patient-derived PDAC tumor and cancer-associated fibroblast cell lines. We applied classification models to independent kinome perturbation and kinase inhibitor cell screen data, and found that the inferred kinotype of a cell has a significant and predictive relationship with cell viability. We further find that models are able to identify a set of kinases whose behavior in response to perturbation drive the majority of viability responses in these cell lines, including the understudied kinases CSNK2A1/3, CAMKK2, and PIP4K2C. We next utilized these models to predict the response of new, clinical kinase inhibitors that were not present in the initial dataset for model devlopment and conducted a validation screen that confirmed the accuracy of the models. These results suggest that characterizing the perturbed state of the human protein kinome provides significant opportunity for better understanding of signaling behavior and downstream cell phenotypes, as well as providing insight into the broader design of potential therapeutic strategies for PDAC.

Evaluating Ovarian Cancer Chemotherapy Response Using Gene Expression Data and Machine Learning.

Background: Ovarian cancer (OC) is the most lethal gynecological cancer in the United States. Among the different types of OC, serous ovarian cancer (SOC) stands out as the most prevalent. Transcriptomics techniques generate extensive gene expression data, yet only a few of these genes are relevant to clinical diagnosis. Methods: Methods for feature selection (FS) address the challenges of high dimensionality in extensive datasets. This study proposes a computational framework that applies FS techniques to identify genes highly associated with platinum-based chemotherapy response on SOC patients. Using SOC datasets from the Gene Expression Omnibus (GEO) database, LASSO and varSelRF FS methods were employed. Machine learning classification algorithms such as random forest (RF) and support vector machine (SVM) were also used to evaluate the performance of the models. Results: The proposed framework has identified biomarkers panels with 9 and 10 genes that are highly correlated with platinum-paclitaxel and platinum-only response in SOC patients, respectively. The predictive models have been trained using the identified gene signatures and accuracy of above 90% was achieved. Conclusions: In this study, we propose that applying multiple feature selection methods not only effectively reduces the number of identified biomarkers, enhancing their biological relevance, but also corroborates the efficacy of drug response prediction models in cancer treatment.

Machine learning-based biomarker screening for acute myeloid leukemia prognosis and therapy from diverse cell-death patterns.

Acute myeloid leukemia (AML) exhibits pronounced heterogeneity and chemotherapy resistance. Aberrant programmed cell death (PCD) implicated in AML pathogenesis suggests PCD-related signatures could serve as biomarkers to predict clinical outcomes and drug response. We utilized 13 PCD pathways, including apoptosis, pyroptosis, ferroptosis, autophagy, necroptosis, cuproptosis, parthanatos, entotic cell death, netotic cell death, lysosome-dependent cell death, alkaliptosis, oxeiptosis, and disulfidptosis to develop predictive models based on 73 machine learning combinations from 10 algorithms. Bulk RNA-sequencing, single-cell RNA-sequencing transcriptomic data, and matched clinicopathological information were obtained from the TCGA-AML, Tyner, and GSE37642-GPL96 cohorts. These datasets were leveraged to construct and validate the models. Additionally, in vitro experiments were conducted to substantiate the bioinformatics findings. The machine learning approach established a 6-gene pan-programmed cell death-related genes index (PPCDI) signature. Validation in two external cohorts showed high PPCDI associated with worse prognosis in AML patients. Incorporating PPCDI with clinical variables, we constructed several robust prognostic nomograms that accurately predicted prognosis of AML patients. Multi-omics analysis integrating bulk and single-cell transcriptomics revealed correlations between PPCDI and immunological features, delineating the immune microenvironment landscape in AML. Patients with high PPCDI exhibited resistance to conventional chemotherapy like doxorubicin but retained sensitivity to dasatinib and methotrexate (FDA-approved drugs for other leukemias), suggesting the potential of PPCDI to guide personalized therapy selection in AML. In summary, we developed a novel PPCDI model through comprehensive analysis of diverse programmed cell death pathways. This PPCDI signature demonstrates great potential in predicting clinical prognosis and drug sensitivity phenotypes in AML patients.

Anoikis-Related Genes Impact Prognosis and Tumor Microenvironment in Bladder Cancer.

Anoikis tolerance is an important biological process of tumor colonization and metastasis outside the primary tumor. Recent research has progressively elucidated the function and underlying mechanisms of anoikis in the metastasis of various solid tumors. Nevertheless, the specific mechanisms of anoikis in bladder cancer and its consequent effects on the tumor immune microenvironment remain ambiguous. In this study, we developed an anoikis score based on five genes (ETV7, NGF, SCD, LAMC1, and CASP6) and categorized subjects into high and low-risk groups using the median score from the TCGA database. Our findings indicate that SCD enhances the proliferation of bladder cancer cells in vitro. Furthermore, integrating the anoikis score with clinicopathological features to construct a prognostic nomogram demonstrated precision in assessing patient outcomes. Immune cell analysis revealed elevated infiltration levels of Treg cells and M2 macrophages in the high anoikis score group, whereas CD8+ T cell levels were reduced. This study highlights the importance of anoikis score in predicting patient prognosis, immune cell infiltration, and drug response, which may provide a treatment modality worth exploring in depth for the study of bladder cancer.

Investigating Topical Steroid Withdrawal Videos on TikTok: Cross-Sectional Analysis of the Top 100 Videos.

BACKGROUND: Social media platforms like TikTok are a very popular source of information, especially for skin diseases. Topical steroid withdrawal (TSW) is a condition that is yet to be fully defined and understood. This did not stop the hashtag #topicalsteroidwithdrawal from amassing more than 600 million views on TikTok. It is of utmost importance to assess the quality and content of TikTok videos on TSW to prevent the spread of misinformation. OBJECTIVE: This study aims to assess the quality and content of the top 100 videos dedicated to the topic of TSW on TikTok. METHODS: This observational study assesses the content and quality of the top 100 videos about TSW on TikTok. A total of 3 independent scoring systems: DISCERN, Journal of the American Medical Association, and Global Quality Scale were used to assess the video quality. The content of the videos was coded by 2 reviewers and analyzed for recurrent themes and topics. RESULTS: This study found that only 10.0% (n=10) of the videos clearly defined what TSW is. Videos were predominantly posted by White, middle-aged, and female creators. Neither cause nor mechanism of the disease were described in the videos. The symptoms suggested itching, peeling, and dryness which resembled the symptoms of atopic dermatitis. The videos fail to mention important information regarding the use of steroids such as the reason it was initially prescribed, the name of the drug, concentration, mechanism of usage, and method of discontinuation. Management techniques varied from hydration methods approved for treatment of atopic dermatitis to treatment options without scientific evidence. Overall, the videos had immense reach with over 200 million views, 45 million likes, 90,000 comments, and 100,000 shares. Video quality was poor with an average DISCERN score of 1.63 (SD 0.56)/5. Video length, total view count, and views/day were all associated with increased quality, indicating that patients were interacting more with higher quality videos. However, videos were created exclusively by personal accounts, highlighting the absence of dermatologists on the platform to discuss this topic. CONCLUSIONS: The videos posted on TikTok are of low quality and lack pertinent information. The content is varied and not consistent. Health care professionals, including dermatologists and residents in the field, need to be more active on the topic, to spread proper information and prevent an increase in steroid phobia. Health care professionals are encouraged to ride the wave and produce high-quality videos discussing what is known about TSW to avoid the spread of misinformation.

DRN-CDR: A cancer drug response prediction model using multi-omics and drug features.

Cancer drug response (CDR) prediction is an important area of research that aims to personalize cancer therapy, optimizing treatment plans for maximum effectiveness while minimizing potential negative effects. Despite the advancements in Deep learning techniques, the effective integration of multi-omics data for drug response prediction remains challenging. In this paper, a regression method using Deep ResNet for CDR (DRN-CDR) prediction is proposed. We aim to explore the potential of considering sole cancer genes in drug response prediction. Here the multi-omics data such as gene expressions, mutation data, and methylation data along with the molecular structural information of drugs were integrated to predict the IC50 values of drugs. Drug features are extracted by employing a Uniform Graph Convolution Network, while Cell line features are extracted using a combination of Convolutional Neural Network and Fully Connected Networks. These features are then concatenated and fed into a deep ResNet for the prediction of IC50 values between Drug - Cell line pairs. The proposed method yielded higher Pearson's correlation coefficient (rp) of 0.7938 with lowest Root Mean Squared Error (RMSE) value of 0.92 when compared with similar methods of tCNNS, MOLI, DeepCDR, TGSA, NIHGCN, DeepTTA, GraTransDRP and TSGCNN. Further, when the model is extended to a classification problem to categorize drugs as sensitive or resistant, we achieved AUC and AUPR measures of 0.7623 and 0.7691, respectively. The drugs such as Tivozanib, SNX-2112, CGP-60474, PHA-665752, Foretinib etc., exhibited low median IC50 values and were found to be effective anti-cancer drugs. The case studies with different TCGA cancer types also revealed the effectiveness of SNX-2112, CGP-60474, Foretinib, Cisplatin, Vinblastine etc. This consistent pattern strongly suggests the effectiveness of the model in predicting CDR.

Association of rare and common genetic variants in MOCOS with inadequate response to allopurinol.

OBJECTIVES: The minor allele of the common rs2231142 ABCG2 variant predicts inadequate response to allopurinol urate lowering therapy. We hypothesize that additional variants in genes encoding urate transporters and allopurinol-to-oxypurinol metabolic enzymes also predict allopurinol response. METHODS: This study included a subset of participants with gout from the Long-term Allopurinol Safety Study Evaluating Outcomes in Gout Patients, whose whole genome was sequenced (n = 563). Good responders had a 4:1 or 5:1 ratio of good (serum urate (SU) <0.36 mmol/l on allopurinol ≤300 mg/day) to poor (SU ≥ 0.36 mmol/l despite allopurinol >300 mg/day) responses over 5-6 timepoints, while inadequate responders had a 1:4 or 1:5 ratio of good to poor responses. Adherence to allopurinol was determined by pill counts, and for a subgroup (n = 303), by plasma oxypurinol >20µmol/l. Using the sequence kernel association test (SKAT) we estimated the combined effect of rare and common variants in urate secretory (ABCC4, ABCC5, ABCG2, SLC17A1, SLC17A3, SLC22A6, SLC22A8) and reuptake genes (SLC2A9, SLC22A11) and in allopurinol-to-oxypurinol metabolic genes (AOX1, MOCOS, XDH) on allopurinol response. RESULTS: There was an association of rare and common variants in the allopurinol-to-oxypurinol gene group (PSKAT-C = 0.019), and in MOCOS, encoding molybdenum cofactor sulphurase, with allopurinol response (PSKAT-C = 0.011). Evidence for genetic association with allopurinol response in the allopurinol-to-oxypurinol gene group (PSKAT-C = 0.002) and MOCOS (PSKAT-C < 0.001) was stronger when adherence to allopurinol therapy was confirmed by plasma oxypurinol. CONCLUSION: We provide evidence for common and rare genetic variation in MOCOS associating with allopurinol response.

An Innovative Multi-Omics Model Integrating Latent Alignment and Attention Mechanism for Drug Response Prediction.

By using omics, we can now examine all components of biological systems simultaneously. Deep learning-based drug prediction methods have shown promise by integrating cancer-related multi-omics data. However, the complex interaction between genes poses challenges in accurately projecting multi-omics data. In this research, we present a predictive model for drug response that incorporates diverse types of omics data, comprising genetic mutation, copy number variation, methylation, and gene expression data. This study proposes latent alignment for information mismatch in integration, which is achieved through an attention module capturing interactions among diverse types of omics data. The latent alignment and attention modules significantly improve predictions, outperforming the baseline model, with MSE = 1.1333, F1-score = 0.5342, and AUROC = 0.5776. High accuracy was achieved in predicting drug responses for piplartine and tenovin-6, while the accuracy was comparatively lower for mitomycin-C and obatoclax. The latent alignment module exclusively outperforms the baseline model, enhancing the MSE by 0.2375, the F1-score by 4.84%, and the AUROC by 6.1%. Similarly, the attention module only improves these metrics by 0.1899, 2.88%, and 2.84%, respectively. In the interpretability case study, panobinostat exhibited the most effective predicted response, with a value of -4.895. We provide reliable insights for drug selection in personalized medicine by identifying crucial genetic factors influencing drug response.

The synergy of artificial intelligence and personalized medicine for the enhanced diagnosis, treatment, and prevention of disease.

INTRODUCTION: The completion of the Human Genome Project in 2003 marked the beginning of a transformative era in medicine. This milestone laid the foundation for personalized medicine, an innovative approach that customizes healthcare treatments. CONTENT: Central to the advancement of personalized medicine is the understanding of genetic variations and their impact on drug responses. The integration of artificial intelligence (AI) into drug response trials has been pivotal in this domain. These technologies excel in handling large-scale genomic datasets and patient histories, significantly improving diagnostic accuracy, disease prediction and drug discovery. They are particularly effective in addressing complex diseases such as cancer and genetic disorders. Furthermore, the advent of wearable technology, when combined with AI, propels personalized medicine forward by offering real-time health monitoring, which is crucial for early disease detection and management. SUMMARY: The integration of AI into personalized medicine represents a significant advancement in healthcare, promising more accurate diagnoses, effective treatment plans and innovative drug discoveries. OUTLOOK: As technology continues to evolve, the role of AI in enhancing personalized medicine and transforming the healthcare landscape is expected to grow exponentially. This synergy between AI and healthcare holds great promise for the future, potentially revolutionizing the way healthcare is delivered and experienced.

18F-FDG PET/CT for early prediction of pathological complete response in breast cancer neoadjuvant therapy: a retrospective analysis.

BACKGROUND: Neoadjuvant treatment has been developed as a systematic approach for patients with early breast cancer and has resulted in improved breast-conserving rate and survival. However, identifying treatment-sensitive patients at the early phase of therapy remains a problem, hampering disease management and raising the possibility of disease progression during treatment. METHODS: In this retrospective analysis, we collected 2-deoxy-2-[F-18] fluoro-d-glucose (18F-FDG) positron emission tomography (PET)/computed tomography (CT) images of primary tumor sites and axillary areas and reciprocal clinical pathological data from 121 patients who underwent neoadjuvant treatment and surgery in our center. The univariate and multivariate logistic regression analyses were performed to investigate features associated with pathological complete response (pCR). An 18F-FDG PET/CT-based prediction model was trained, and the performance was evaluated by receiver operating characteristic curves (ROC). RESULTS: The maximum standard uptake values (SUVmax) of 18F-FDG PET/CT were a powerful indicator of tumor status. The SUVmax values of axillary areas were closely related to metastatic lymph node counts (R = 0.62). Moreover, the early SUVmax reduction rates (between baseline and second cycle of neoadjuvant treatment) were statistically different between pCR and non-pCR patients. The early SUVmax reduction rates-based model showed great ability to predict pCR (AUC = 0.89), with all molecular subtypes (HR+HER2-, HR+HER2+, HR-HER2+, and HR-HER2-) considered. CONCLUSION: Our research proved that the SUVmax reduction rate of 18F-FDG PET/CT contributed to the early prediction of pCR, providing rationales for utilizing PET/CT in NAT in the future.

Mitochondrial Protein Density, Biomass, and Bioenergetics as Predictors for the Efficacy of Glioma Treatments.

The metabolism of glioma cells exhibits significant heterogeneity and is partially responsible for treatment outcomes. Given this variability, we hypothesized that the effectiveness of treatments targeting various metabolic pathways depends on the bioenergetic profiles and mitochondrial status of glioma cells. To this end, we analyzed mitochondrial biomass, mitochondrial protein density, oxidative phosphorylation (OXPHOS), and glycolysis in a panel of eight glioma cell lines. Our findings revealed considerable variability: mitochondrial biomass varied by up to 3.2-fold, the density of mitochondrial proteins by up to 2.1-fold, and OXPHOS levels by up to 7.3-fold across the cell lines. Subsequently, we stratified glioma cell lines based on their mitochondrial status, OXPHOS, and bioenergetic fitness. Following this stratification, we utilized 16 compounds targeting key bioenergetic, mitochondrial, and related pathways to analyze the associations between induced changes in cell numbers, proliferation, and apoptosis with respect to their steady-state mitochondrial and bioenergetic metrics. Remarkably, a significant fraction of the treatments showed strong correlations with mitochondrial biomass and the density of mitochondrial proteins, suggesting that mitochondrial status may reflect glioma cell sensitivity to specific treatments. Overall, our results indicate that mitochondrial status and bioenergetics are linked to the efficacy of treatments targeting metabolic pathways in glioma.

ADAR-Mediated A>I(G) RNA Editing in the Genotoxic Drug Response of Breast Cancer.

Epitranscriptomics is a field that delves into post-transcriptional changes. Among these modifications, the conversion of adenosine to inosine, traduced as guanosine (A>I(G)), is one of the known RNA-editing mechanisms, catalyzed by ADARs. This type of RNA editing is the most common type of editing in mammals and contributes to biological diversity. Disruption in the A>I(G) RNA-editing balance has been linked to diseases, including several types of cancer. Drug resistance in patients with cancer represents a significant public health concern, contributing to increased mortality rates resulting from therapy non-responsiveness and disease progression, representing the greatest challenge for researchers in this field. The A>I(G) RNA editing is involved in several mechanisms over the immunotherapy and genotoxic drug response and drug resistance. This review investigates the relationship between ADAR1 and specific A>I(G) RNA-edited sites, focusing particularly on breast cancer, and the impact of these sites on DNA damage repair and the immune response over anti-cancer therapy. We address the underlying mechanisms, bioinformatics, and in vitro strategies for the identification and validation of A>I(G) RNA-edited sites. We gathered databases related to A>I(G) RNA editing and cancer and discussed the potential clinical and research implications of understanding A>I(G) RNA-editing patterns. Understanding the intricate role of ADAR1-mediated A>I(G) RNA editing in breast cancer holds significant promise for the development of personalized treatment approaches tailored to individual patients' A>I(G) RNA-editing profiles.

Genomic and single-cell characterization of patient-derived tumor organoid models of head and neck squamous cell carcinoma.

Head and Neck Squamous Cell Carcinoma (HNSCC) remains a significant health burden due to tumor heterogeneity and treatment resistance, emphasizing the need for improved biological understanding and tailored therapies. This study enrolled 31 HNSCC patients for the establishment of patient-derived tumor organoids (PDOs), which faithfully maintained genomic features and histopathological traits of primary tumors. Long-term culture preserved key characteristics, affirming PDOs as robust representative models. PDOs demonstrated predictive capability for cisplatin treatment responses, correlating ex vivo drug sensitivity with patient outcomes. Bulk and single-cell RNA sequencing unveiled molecular subtypes and intratumor heterogeneity (ITH) in PDOs, paralleling patient tumors. Notably, a hybrid epithelial-mesenchymal transition (hEMT)-like ITH program is associated with cisplatin resistance and poor patient survival. Functional analyses identified amphiregulin (AREG) as a potential regulator of the hybrid epithelial/mesenchymal state. Moreover, AREG contributes to cisplatin resistance via EGFR pathway activation, corroborated by clinical samples. In summary, HNSCC PDOs serve as reliable and versatile models, offer predictive insights into ITH programs and treatment responses, and uncover potential therapeutic targets for personalized medicine.

Comparative analyses of gene networks mediating cancer metastatic potentials across lineage types.

Studies have identified genes and molecular pathways regulating cancer metastasis. However, it remains largely unknown whether metastatic potentials of cancer cells from different lineage types are driven by the same or different gene networks. Here, we aim to address this question through integrative analyses of 493 human cancer cells' transcriptomic profiles and their metastatic potentials in vivo. Using an unsupervised approach and considering both gene coexpression and protein-protein interaction networks, we identify different gene networks associated with various biological pathways (i.e. inflammation, cell cycle, and RNA translation), the expression of which are correlated with metastatic potentials across subsets of lineage types. By developing a regularized random forest regression model, we show that the combination of the gene module features expressed in the native cancer cells can predict their metastatic potentials with an overall Pearson correlation coefficient of 0.90. By analyzing transcriptomic profile data from cancer patients, we show that these networks are conserved in vivo and contribute to cancer aggressiveness. The intrinsic expression levels of these networks are correlated with drug sensitivity. Altogether, our study provides novel comparative insights into cancer cells' intrinsic gene networks mediating metastatic potentials across different lineage types, and our results can potentially be useful for designing personalized treatments for metastatic cancers.