Harnessing Connectivity in a Large-Scale Small-Molecule Sensitivity Dataset

CurveCurator: a recalibrated F-statistic to assess, classify, and explore significance of dose-response curves

Dose-response curves are key metrics in pharmacology and biology to assess phenotypic or molecular actions of bioactive compounds in a quantitative fashion. Yet, it is often unclear whether or not a measured response significantly differs from a curve without regulation, particularly in high-throughput applications or unstable assays. Treating potency and effect size estimates from random and true curves with the same level of confidence can lead to incorrect hypotheses and issues in training machine learning models. Here, we present CurveCurator, an open-source software that provides reliable dose-response characteristics by computing p-values and false discovery rates based on a recalibrated F-statistic and a target-decoy procedure that considers dataset-specific effect size distributions. The application of CurveCurator to three large-scale datasets enables a systematic drug mode of action analysis and demonstrates its scalable utility across several application areas, facilitated by a performant, interactive dashboard for fast data exploration.

Targeted Investigational Oncology Agents in the NCI-60: A Phenotypic Systems-based Resource

The NCI-60 human tumor cell line panel has proved to be a useful tool for the global cancer research community in the search for novel chemotherapeutics. The publicly available cell line characterization and compound screening data from the NCI-60 assay have significantly contributed to the understanding of cellular mechanisms targeted by new oncology agents. Signature sensitivity/resistance patterns generated for a given chemotherapeutic agent against the NCI-60 panel have long served as fingerprint presentations that encompass target information and the mechanism of action associated with the tested agent. We report the establishment of a new public NCI-60 resource based on the cell line screening of a large and growing set of 175 FDA-approved oncology drugs (AOD) plus >825 clinical and investigational oncology agents (IOA), representing a diverse set (>250) of therapeutic targets and mechanisms. This data resource is available to the public (https://ioa.cancer.gov) and includes the raw data from the screening of the IOA and AOD collection along with an extensive set of visualization and analysis tools to allow for comparative study of individual test compounds and multiple compound sets.

Inferring therapeutic vulnerability within tumors through integration of pan-cancer cell line and single-cell transcriptomic profiles

Single-cell RNA sequencing greatly advanced our understanding of intratumoral heterogeneity through identifying tumor subpopulations with distinct biologies. However, translating biological differences into treatment strategies is challenging, as we still lack tools to facilitate efficient drug discovery that tackles heterogeneous tumors. One key component of such approaches tackles accurate prediction of drug response at the single-cell level to offer therapeutic options to specific cell subpopulations. Here, we present a transparent computational framework (nicknamed scIDUC) to predict therapeutic efficacies on an individual-cell basis by integrating single-cell transcriptomic profiles with large, data-rich pan-cancer cell line screening datasets. Our method achieves high accuracy, with predicted sensitivities easily able to separate cells into their true cellular drug resistance status as measured by effect size (Cohen's d > 1.0). More importantly, we examine our method's utility with three distinct prospective tests covering different diseases (rhabdomyosarcoma, pancreatic ductal adenocarcinoma, and castration-resistant prostate cancer), and in each our predicted results are accurate and mirrored biological expectations. In the first two, we identified drugs for cell subpopulations that are resistant to standard-of-care (SOC) therapies due to intrinsic resistance or effects of tumor microenvironments. Our results showed high consistency with experimental findings from the original studies. In the third test, we generated SOC therapy resistant cell lines, used scIDUC to identify efficacious drugs for the resistant line, and validated the predictions with in-vitro experiments. Together, scIDUC quickly translates scRNA-seq data into drug response for individual cells, displaying the potential as a first-line tool for nuanced and heterogeneity-aware drug discovery.

Pharmacological depletion of RNA splicing factor RBM39 by indisulam synergizes with PARP inhibitors in high-grade serous ovarian carcinoma

Ovarian high-grade serous carcinoma (HGSC) is the most common subtype of ovarian cancer with limited therapeutic options and a poor prognosis. In recent years, poly-ADP ribose polymerase (PARP) inhibitors have demonstrated significant clinical benefits, especially in patients with BRCA1/2 mutations. However, acquired drug resistance and relapse is a major challenge. Indisulam (E7070) has been identified as a molecular glue that brings together splicing factor RBM39 and DCAF15 E3 ubiquitin ligase resulting in polyubiquitination, degradation, and subsequent RNA splicing defects. In this work, we demonstrate that the loss of RBM39 induces splicing defects in key DNA damage repair genes in ovarian cancer, leading to increased sensitivity to cisplatin and various PARP inhibitors. The addition of indisulam also improved olaparib response in mice bearing PARP inhibitor-resistant tumors. These findings demonstrate that combining RBM39 degraders and PARP inhibitors is a promising therapeutic approach to improve PARP inhibitor response in ovarian HGSC.

Pan-cancer analysis revealing that PTPN2 is an indicator of risk stratification for acute myeloid leukemia

The non-receptor protein tyrosine phosphatases gene family (PTPNs) is involved in the tumorigenesis and development of many cancers, but the role of PTPNs in acute myeloid leukemia (AML) remains unclear. After a comprehensive evaluation on the expression patterns and immunological effects of PTPNs using a pan-cancer analysis based on RNA sequencing data obtained from The Cancer Genome Atlas, the most valuable gene PTPN2 was discovered. Further investigation of the expression patterns of PTPN2 in different tissues and cells showed a robust correlation with AML. PTPN2 was then systematically correlated with immunological signatures in the AML tumor microenvironment and its differential expression was verified using clinical samples. In addition, a prediction model, being validated and compared with other models, was developed in our research. The systematic analysis of PTPN family reveals that the effect of PTPNs on cancer may be correlated to mediating cell cycle-related pathways. It was then found that PTPN2 was highly expressed in hematologic diseases and bone marrow tissues, and its differential expression in AML patients and normal humans was verified by clinical samples. Based on its correlation with immune infiltrates, immunomodulators, and immune checkpoint, PTPN2 was found to be a reliable biomarker in the immunotherapy cohort and a prognostic predictor of AML. And PTPN2'riskscore can accurately predict the prognosis and response of cancer immunotherapy. These findings revealed the correlation between PTPNs and immunophenotype, which may be related to cell cycle. PTPN2 was differentially expressed between clinical AML patients and normal people. It is a diagnostic biomarker and potentially therapeutic target, providing targeted guidance for clinical treatment.

Chromatin accessibility landscape of relapsed pediatric B-lineage acute lymphoblastic leukemia

For around half of the pediatric B-lineage acute lymphoblastic leukemia (B-ALL) patients, the molecular mechanism of relapse remains unclear. To fill this gap in knowledge, here we characterize the chromatin accessibility landscape in pediatric relapsed B-ALL. We observe rewired accessible chromatin regions (ACRs) associated with transcription dysregulation in leukemia cells as compared with normal B-cell progenitors. We show that over a quarter of the ACRs in B-ALL are in quiescent regions with high heterogeneity among B-ALLs. We identify subtype-specific and allele-imbalanced chromatin accessibility by integrating multi-omics data. By characterizing the differential ACRs between diagnosis and relapse in B-ALL, we identify alterations in chromatin accessibility during drug treatment. Further analysis of ACRs associated with relapse free survival leads to the identification of a subgroup of B-ALL which show early relapse. These data provide an advanced and integrative portrait of the importance of chromatin accessibility alterations in tumorigenesis and drug responses.

Computational identification and clinical validation of a novel risk signature based on coagulation-related lncRNAs for predicting prognosis, immunotherapy response, and chemosensitivity in colorectal cancer patients

Background

Coagulation is critically involved in the tumor microenvironment, cancer progression, and prognosis assessment. Nevertheless, the roles of coagulation-related long noncoding RNAs (CRLs) in colorectal cancer (CRC) remain unclear. In this study, an integrated computational framework was constructed to develop a novel coagulation-related lncRNA signature (CRLncSig) to stratify the prognosis of CRC patients, predict response to immunotherapy and chemotherapy in CRC, and explore the potential molecular mechanism.

Methods

CRC samples from The Cancer Genome Atlas (TCGA) were used as the training set, while the substantial bulk or single-cell RNA transcriptomics from Gene Expression Omnibus (GEO) datasets and real-time quantitative PCR (RT-qPCR) data from CRC cell lines and paired frozen tissues were used for validation. We performed unsupervised consensus clustering of CRLs to classify patients into distinct molecular subtypes. We then used stepwise regression to establish the CRLncSig risk model, which stratified patients into high- and low-risk groups. Subsequently, diversified bioinformatics algorithms were used to explore prognosis, biological pathway alteration, immune microenvironment, immunotherapy response, and drug sensitivity across patient subgroups. In addition, weighted gene coexpression network analysis was used to construct an lncRNA-miRNA-mRNA competitive endogenous network. Expression levels of CRLncSig, immune checkpoints, and immunosuppressors were determined using RT-qPCR.

Results

We identified two coagulation subclusters and constructed a risk score model using CRLncSig in CRC, where the patients in cluster 2 and the low-risk group had a better prognosis. The cluster and CRLncSig were confirmed as the independent risk factors, and a CRLncSig-based nomogram exhibited a robust prognostic performance. Notably, the cluster and CRLncSig were identified as the indicators of immune cell infiltration, immunoreactivity phenotype, and immunotherapy efficiency. In addition, we identified a new endogenous network of competing CRLs with microRNA/mRNA, which will provide a foundation for future mechanistic studies of CRLs in the malignant progression of CRC. Moreover, CRLncSig strongly correlated with drug susceptibility.

Conclusion

We developed a reliable CRLncSig to predict the prognosis, immune landscape, immunotherapy response, and drug sensitivity in patients with CRC, which might facilitate optimizing risk stratification, guiding the applications of immunotherapy, and individualized treatments for CRC.

Reassessing pharmacogenomic cell sensitivity with multilevel statistical models

Pharmacogenomic experiments allow for the systematic testing of drugs, at varying dosage concentrations, to study how genomic markers correlate with cell sensitivity to treatment. The first step in the analysis is to quantify the response of cell lines to variable dosage concentrations of the drugs being tested. The signal to noise in these measurements can be low due to biological and experimental variability. However, the increasing availability of pharmacogenomic studies provides replicated data sets that can be leveraged to gain power. To do this, we formulate a hierarchical mixture model to estimate the drug-specific mixture distributions for estimating cell sensitivity and for assessing drug effect type as either broad or targeted effect. We use this formulation to propose a unified approach that can yield posterior probability of a cell being susceptible to a drug conditional on being a targeted effect or relative effect sizes conditioned on the cell being broad. We demonstrate the usefulness of our approach via case studies. First, we assess pairwise agreements for cell lines/drugs within the intersection of two data sets and confirm the moderate pairwise agreement between many publicly available pharmacogenomic data sets. We then present an analysis that identifies sensitivity to the drug crizotinib for cells harboring EML4-ALK or NPM1-ALK gene fusions, as well as significantly down-regulated cell-matrix pathways associated with crizotinib sensitivity.

Cross-Talks between RKIP and YY1 through a Multilevel Bioinformatics Pan-Cancer Analysis

Recent studies suggest that PEBP1 (also known as RKIP) and YY1, despite having distinct molecular functions, may interact and mutually influence each other's activity. They exhibit reciprocal control over each other's expression through regulatory loops, prompting the hypothesis that their interplay could be pivotal in cancer advancement and resistance to drugs. To delve into this interplay's functional characteristics, we conducted a comprehensive analysis using bioinformatics tools across a range of cancers. Our results confirm the association between elevated YY1 mRNA levels and varying survival outcomes in diverse tumors. Furthermore, we observed differing degrees of inhibitory or activating effects of these two genes in apoptosis, cell cycle, DNA damage, and other cancer pathways, along with correlations between their mRNA expression and immune infiltration. Additionally, YY1/PEBP1 expression and methylation displayed connections with genomic alterations across different cancer types. Notably, we uncovered links between the two genes and different indicators of immunosuppression, such as immune checkpoint blockade response and T-cell dysfunction/exclusion levels, across different patient groups. Overall, our findings underscore the significant role of the interplay between YY1 and PEBP1 in cancer progression, influencing genomic changes, tumor immunity, or the tumor microenvironment. Additionally, these two gene products appear to impact the sensitivity of anticancer drugs, opening new avenues for cancer therapy.

Targeting pan-essential pathways in cancer with cytotoxic chemotherapy: challenges and opportunities

Cytotoxic chemotherapy remains a key modality in cancer treatment. These therapies, successfully used for decades, continue to transform the lives of cancer patients daily. With the high attrition rate of current oncology drug development, combined with the knowledge that most new therapies do not displace standard-of-care treatments and that many healthcare systems cannot afford these new therapies; cytotoxic chemotherapies will remain an important component of cancer therapy for many years to come. The clinical value of these therapies is often under-appreciated within the pre-clinical cancer research community, where this diverse class of agents are often grouped together as non-specific cellular poisons killing tumor cells based solely upon proliferation rate; however, this is inaccurate. This review article seeks to reaffirm the importance of focusing research efforts upon improving our basic understanding of how these drugs work, discussing their ability to target pan-essential pathways in cancer cells, the relationship of this to the chemotherapeutic window, and highlighting basic science approaches that can be employed towards refining their use.

MMD collaborates with ACSL4 and MBOAT7 to promote polyunsaturated phosphatidylinositol remodeling and susceptibility to ferroptosis

Ferroptosis is a form of regulated cell death with roles in degenerative diseases and cancer. Excessive iron-catalyzed peroxidation of membrane phospholipids, especially those containing the polyunsaturated fatty acid arachidonic acid (AA), is central in driving ferroptosis. Here, we reveal that an understudied Golgi-resident scaffold protein, MMD, promotes susceptibility to ferroptosis in ovarian and renal carcinoma cells in an ACSL4- and MBOAT7-dependent manner. Mechanistically, MMD physically interacts with both ACSL4 and MBOAT7, two enzymes that catalyze sequential steps to incorporate AA in phosphatidylinositol (PI) lipids. Thus, MMD increases the flux of AA into PI, resulting in heightened cellular levels of AA-PI and other AA-containing phospholipid species. This molecular mechanism points to a pro-ferroptotic role for MBOAT7 and AA-PI, with potential therapeutic implications, and reveals that MMD is an important regulator of cellular lipid metabolism.

Integrative analysis of cancer dependency data and comprehensive phosphoproteomics data revealed the EPHA2-PARD3 axis as a cancer vulnerability in KRAS-mutant colorectal cancer

Colorectal cancer (CRC), a common malignant tumour of the gastrointestinal tract, is a life-threatening cancer worldwide. Mutations in KRAS and BRAF, the major driver mutation subtypes in CRC, activate the RAS pathway, contribute to tumorigenesis in CRC and are being investigated as potential therapeutic targets. Despite recent advances in clinical trials targeting KRASG12C or RAS downstream signalling molecules for KRAS-mutant CRC, there is a lack of effective therapeutic interventions. Therefore, understanding the unique molecular characteristics of KRAS-mutant CRC is essential for identifying molecular targets and developing novel therapeutic interventions. We obtained in-depth proteomics and phosphoproteomics quantitative data for over 7900 proteins and 38 700 phosphorylation sites in cells from 35 CRC cell lines and performed informatic analyses, including proteomics-based coexpression analysis and correlation analysis between phosphoproteomics data and cancer dependency scores of the corresponding phosphoproteins. Our results revealed novel dysregulated protein-protein associations enriched specifically in KRAS-mutant cells. Our phosphoproteomics analysis revealed activation of EPHA2 kinase and downstream tight junction signalling in KRAS-mutant cells. Furthermore, the results implicate the phosphorylation site Y378 in the tight junction protein PARD3 as a cancer vulnerability in KRAS-mutant cells. Together, our large-scale phosphoproteomics and proteomics data across 35 steady-state CRC cell lines represent a valuable resource for understanding the molecular characteristics of oncogenic mutations. Our approach to predicting cancer dependency from phosphoproteomics data identified the EPHA2-PARD3 axis as a cancer vulnerability in KRAS-mutant CRC.

Simplicity: web-based visualization and analysis of high-throughput cancer cell line screens

High-throughput drug screens are a powerful tool for cancer drug development. However, the results of such screens are often made available only as raw data, which is intractable for researchers without informatic skills, or as highly processed summary statistics, which can lack essential information for translating screening results into clinically meaningful discoveries. To improve the usability of these datasets, we developed Simplicity, a robust and user-friendly web interface for visualizing, exploring, and summarizing raw and processed data from high-throughput drug screens. Importantly, Simplicity allows for easy recalculation of summary statistics at user-defined drug concentrations. This allows Simplicity's outputs to be used with methods that rely on statistics being calculated at clinically relevant doses. Simplicity can be freely accessed at https://oncotherapyinformatics.org/simplicity/.

Pan-Cancer Pharmacogenomic Analysis of Patient-Derived Tumor Cells Using Clinically Relevant Drug Exposures

As a result of tumor heterogeneity and solid cancers harboring multiple molecular defects, precision medicine platforms in oncology are most effective when both genetic and pharmacologic determinants of a tumor are evaluated. Expandable patient-derived xenograft (PDX) mouse tumor and corresponding PDX culture (PDXC) models recapitulate many of the biological and genetic characteristics of the original patient tumor, allowing for a comprehensive pharmacogenomic analysis. Here, the somatic mutations of 23 matched patient tumor and PDX samples encompassing four cancers were first evaluated using next-generation sequencing (NGS). 19 antitumor agents were evaluated across 78 patient-derived tumor cultures using clinically relevant drug exposures. A binarization threshold sensitivity classification determined in culture (PDXC) was used to identify tumors that best respond to drug in vivo (PDX). Using this sensitivity classification, logic models of DNA mutations were developed for 19 antitumor agents to predict drug response. We determined that the concordance of somatic mutations across patient and corresponding PDX samples increased as variant allele frequency increased. Notable individual PDXC responses to specific drugs, as well as lineage-specific drug responses were identified. Robust responses identified in PDXC were recapitulated in vivo in PDX-bearing mice and logic modeling determined somatic gene mutation(s) defining response to specific antitumor agents. In conclusion, combining NGS of primary patient tumors, high-throughput drug screen using clinically relevant doses, and logic modeling, can provide a platform for understanding response to therapeutic drugs targeting cancer.

BQsupports: systematic assessment of the support and novelty of new biomedical associations

MOTIVATION

Living a Big Data era in Biomedicine, there is an unmet need to systematically assess experimental observations in the context of available information. This assessment would offer a means for a comprehensive and robust validation of biomedical data results and provide an initial estimate of the potential novelty of the findings.

RESULTS

Here we present BQsupports, a web-based tool built upon the Bioteque biomedical descriptors that systematically analyzes and quantifies the current support to a given set of observations. The tool relies on over 1000 distinct types of biomedical descriptors, covering over 11 different biological and chemical entities, including genes, cell lines, diseases, and small molecules. By exploring hundreds of descriptors, BQsupports provide support scores for each observation across a wide variety of biomedical contexts. These scores are then aggregated to summarize the biomedical support of the assessed dataset as a whole. Finally, the BQsupports also suggests predictive features of the given dataset, which can be exploited in downstream machine learning applications.

AVAILABILITY AND IMPLEMENTATION

The web application and underlying data are available online (https://bqsupports.irbbarcelona.org).

Partial gene suppression improves identification of cancer vulnerabilities when CRISPR-Cas9 knockout is pan-lethal

BACKGROUND

Hundreds of functional genomic screens have been performed across a diverse set of cancer contexts, as part of efforts such as the Cancer Dependency Map, to identify gene dependencies-genes whose loss of function reduces cell viability or fitness. Recently, large-scale screening efforts have shifted from RNAi to CRISPR-Cas9, due to superior efficacy and specificity. However, many effective oncology drugs only partially inhibit their protein targets, leading us to question whether partial suppression of genes using RNAi could reveal cancer vulnerabilities that are missed by complete knockout using CRISPR-Cas9. Here, we compare CRISPR-Cas9 and RNAi dependency profiles of genes across approximately 400 matched cancer cell lines.

RESULTS

We find that CRISPR screens accurately identify more gene dependencies per cell line, but the majority of each cell line's dependencies are part of a set of 1867 genes that are shared dependencies across the entire collection (pan-lethals). While RNAi knockdown of about 30% of these genes is also pan-lethal, approximately 50% have selective dependency patterns across cell lines, suggesting they could still be cancer vulnerabilities. The accuracy of the unique RNAi selectivity is supported by associations to multi-omics profiles, drug sensitivity, and other expected co-dependencies.

CONCLUSIONS

Incorporating RNAi data for genes that are pan-lethal knockouts facilitates the discovery of a wider range of gene targets than could be detected using the CRISPR dataset alone. This can aid in the interpretation of contrasting results obtained from CRISPR and RNAi screens and reinforce the importance of partial gene suppression methods in building a cancer dependency map.

The pharmacoepigenomic landscape of cancer cell lines reveals the epigenetic component of drug sensitivity

Aberrant DNA methylation accompanies genetic alterations during oncogenesis and tumour homeostasis and contributes to the transcriptional deregulation of key signalling pathways in cancer. Despite increasing efforts in DNA methylation profiling of cancer patients, there is still a lack of epigenetic biomarkers to predict treatment efficacy. To address this, we analyse 721 cancer cell lines across 22 cancer types treated with 453 anti-cancer compounds. We systematically detect the predictive component of DNA methylation in the context of transcriptional and mutational patterns, i.e., in total 19 DNA methylation biomarkers across 17 drugs and five cancer types. DNA methylation constitutes drug sensitivity biomarkers by mediating the expression of proximal genes, thereby enhancing biological signals across multi-omics data modalities. Our method reproduces anticipated associations, and in addition, we find that the NEK9 promoter hypermethylation may confer sensitivity to the NEDD8-activating enzyme (NAE) inhibitor pevonedistat in melanoma through downregulation of NEK9. In summary, we envision that epigenomics will refine existing patient stratification, thus empowering the next generation of precision oncology.

Machine learning-driven exploration of drug therapies for triple-negative breast cancer treatment

Breast cancer is the second leading cause of cancer death in women among all cancer types. It is highly heterogeneous in nature, which means that the tumors have different morphologies and there is heterogeneity even among people who have the same type of tumor. Several staging and classifying systems have been developed due to the variability of different types of breast cancer. Due to high heterogeneity, personalized treatment has become a new strategy. Out of all breast cancer subtypes, triple-negative breast cancer (TNBC) comprises ∼10%-15%. TNBC refers to the subtype of breast cancer where cells do not express estrogen receptors, progesterone receptors, or human epidermal growth factor receptors (ERs, PRs, and HERs). Tumors in TNBC have a diverse set of genetic markers and prognostic indicators. We scanned the Cancer Cell Line Encyclopedia (CCLE) and Genomics of Drug Sensitivity in Cancer (GDSC) databases for potential drugs using human breast cancer cell lines and drug sensitivity data. Three different machine-learning approaches were used to evaluate the prediction of six effective drugs against the TNBC cell lines. The top biomarkers were then shortlisted on the basis of their involvement in breast cancer and further subjected to testing for radion resistance using data from the Cleveland database. It was observed that Panobinostat, PLX4720, Lapatinib, Nilotinib, Selumetinib, and Tanespimycin were six effective drugs against the TNBC cell lines. We could identify potential derivates that may be used against approved drugs. Only one biomarker (SETD7) was sensitive to all six drugs on the shortlist, while two others (SRARP and YIPF5) were sensitive to both radiation and drugs. Furthermore, we did not find any radioresistance markers for the TNBC. The proposed biomarkers and drug sensitivity analysis will provide potential candidates for future clinical investigation.

XMR: an explainable multimodal neural network for drug response prediction

Introduction: Existing large-scale preclinical cancer drug response databases provide us with a great opportunity to identify and predict potentially effective drugs to combat cancers. Deep learning models built on these databases have been developed and applied to tackle the cancer drug-response prediction task. Their prediction has been demonstrated to significantly outperform traditional machine learning methods. However, due to the "black box" characteristic, biologically faithful explanations are hardly derived from these deep learning models. Interpretable deep learning models that rely on visible neural networks (VNNs) have been proposed to provide biological justification for the predicted outcomes. However, their performance does not meet the expectation to be applied in clinical practice. Methods: In this paper, we develop an XMR model, an eXplainable Multimodal neural network for drug Response prediction. XMR is a new compact multimodal neural network consisting of two sub-networks: a visible neural network for learning genomic features and a graph neural network (GNN) for learning drugs' structural features. Both sub-networks are integrated into a multimodal fusion layer to model the drug response for the given gene mutations and the drug's molecular structures. Furthermore, a pruning approach is applied to provide better interpretations of the XMR model. We use five pathway hierarchies (cell cycle, DNA repair, diseases, signal transduction, and metabolism), which are obtained from the Reactome Pathway Database, as the architecture of VNN for our XMR model to predict drug responses of triple negative breast cancer. Results: We find that our model outperforms other state-of-the-art interpretable deep learning models in terms of predictive performance. In addition, our model can provide biological insights into explaining drug responses for triple-negative breast cancer. Discussion: Overall, combining both VNN and GNN in a multimodal fusion layer, XMR captures key genomic and molecular features and offers reasonable interpretability in biology, thereby better predicting drug responses in cancer patients. Our model would also benefit personalized cancer therapy in the future.

Donafenib and GSK-J4 Synergistically Induce Ferroptosis in Liver Cancer by Upregulating HMOX1 Expression

Hepatocellular carcinoma (HCC) is one of the most lethal cancers worldwide. Donafenib is a multi-receptor tyrosine kinase inhibitor approved for the treatment of patients with advanced HCC, but its clinical effect is very limited. Here, through integrated screening of a small-molecule inhibitor library and a druggable CRISPR library, that GSK-J4 is synthetically lethal with donafenib in liver cancer is shown. This synergistic lethality is validated in multiple HCC models, including xenograft, orthotopically induced HCC, patient-derived xenograft, and organoid models. Furthermore, co-treatment with donafenib and GSK-J4 resulted in cell death mainly via ferroptosis. Mechanistically, through integrated RNA sequencing (RNA-seq) and assay for transposase-accessible chromatin with high throughput sequencing (ATAC-seq) analyses, that donafenib and GSK-J4 synergistically promoted the expression of HMOX1 and increased the intracellular Fe2+ level is found, eventually leading to ferroptosis. Additionally, through cleavage under targets & tagmentation followed by sequencing (CUT&Tag-seq), it is found that the enhancer regions upstream of HMOX1 promoter significantly increased under donafenib and GSK-J4 co-treatment. A chromosome conformation capture assay confirmed that the increased expression of HMOX1 is caused by the significantly enhanced interaction between the promoter and upstream enhancer under dual-drug combination. Taken together, this study elucidates a new synergistic lethal interaction in liver cancer.

Machine learning model for anti-cancer drug combinations: Analysis, prediction, and validation

Drug combination therapy is a highly effective approach for enhancing the therapeutic efficacy of anti-cancer drugs and overcoming drug resistance. However, the innumerable possible drug combinations make it impractical to screen all synergistic drug pairs. Moreover, biological insights into synergistic drug pairs are still lacking. To address this challenge, we systematically analyzed drug combination datasets curated from multiple databases to identify drug pairs more likely to show synergy. We classified drug pairs based on their MoA and discovered that 110 MoA pairs were significantly enriched in synergy in at least one type of cancer. To improve the accuracy of predicting synergistic effects of drug pairs, we developed a suite of machine learning models that achieve better predictive performance. Unlike most previous methods that were rarely validated by wet-lab experiments, our models were validated using two-dimensional cell lines and three-dimensional tumor slice culture (3D-TSC) models, implying their practical utility. Our prediction and validation results indicated that the combination of the RTK inhibitors Lapatinib and Pazopanib exhibited a strong therapeutic effect in breast cancer by blocking the downstream PI3K/AKT/mTOR signaling pathway. Furthermore, we incorporated molecular features to identify potential biomarkers for synergistic drug pairs, and almost all potential biomarkers found connections between drug targets and corresponding molecular features using protein-protein interaction network. Overall, this study provides valuable insights to complement and guide rational efforts to develop drug combination treatments.

Predicting anticancer drug sensitivity on distributed data sources using federated deep learning

Drug sensitivity prediction plays a crucial role in precision cancer therapy. Collaboration among medical institutions can lead to better performance in drug sensitivity prediction. However, patient privacy and data protection regulation remain a severe impediment to centralized prediction studies. For the first time, we proposed a federated drug sensitivity prediction model with high generalization, combining distributed data sources while protecting private data. Cell lines are first classified into three categories using the waterfall method. Focal loss for solving class imbalance is then embedded into the horizontal federated deep learning framework, i.e., HFDL-fl is presented. Applying HFDL-fl to homogeneous and heterogeneous data, we obtained HFDL-Cross and HFDL-Within. Our comprehensive experiments demonstrated that (i) collaboration by HFDL-fl outperforms private model on local data, (ii) focal loss function can effectively improve model performance to classify cell lines in sensitive and resistant categories, and (iii) HFDL-fl is not significantly affected by data heterogeneity. To summarize, HFDL-fl provides a valuable solution to break down the barriers between medical institutions for privacy-preserving drug sensitivity prediction and therefore facilitates the development of cancer precision medicine and other privacy-related biomedical research.

Chemogenomic library design strategies for precision oncology, applied to phenotypic profiling of glioblastoma patient cells

Designing a targeted screening library of bioactive small molecules is a challenging task since most compounds modulate their effects through multiple protein targets with varying degrees of potency and selectivity. We implemented analytic procedures for designing anticancer compound libraries adjusted for library size, cellular activity, chemical diversity and availability, and target selectivity. The resulting compound collections cover a wide range of protein targets and biological pathways implicated in various cancers, making them widely applicable to precision oncology. We characterized the compound and target spaces of the virtual libraries, in comparison with a minimal screening library of 1,211 compounds for targeting 1,386 anticancer proteins. In a pilot screening study, we identified patient-specific vulnerabilities by imaging glioma stem cells from patients with glioblastoma (GBM), using a physical library of 789 compounds that cover 1,320 of the anticancer targets. The cell survival profiling revealed highly heterogeneous phenotypic responses across the patients and GBM subtypes.

Machine learning-based risk model incorporating tumor immune and stromal contexture predicts cancer prognosis and immunotherapy efficacy

The immune and stromal contexture within the tumor microenvironment (TME) interact with cancer cells and jointly determine disease process and therapeutic response. We aimed at developing a risk scoring model based on TME-related genes of squamous cell lung cancer to predict patient prognosis and immunotherapeutic response. TME-related genes were identified through exploring genes that correlated with immune scores and stromal scores. LASSO-Cox regression model was used to establish the TME-related risk scoring (TMErisk) model. A TMErisk model containing six genes was established. High TMErisk correlated with unfavorable OS in LUSC patients and this association was validated in multiple NSCLC datasets. Genes involved in pathways associated with immunosuppressive microenvironment were enriched in the high TMErisk group. Tumors with high TMErisk showed elevated infiltration of immunosuppressive cells. High TMErisk predicted worse immunotherapeutic response and prognosis across multiple carcinomas. TMErisk model could serve as a robust biomarker for predicting OS and immunotherapeutic response.

Inhibition of CDK9 exhibits anticancer activity in hepatocellular carcinoma cells via targeting ribonucleotide reductase

Cyclin-dependent kinase 9 (CDK9) inhibitors are a novel category of anticancer treatment for cancers. However, their effects on hepatocellular carcinoma (HCC) are rarely investigated. Human ribonucleotide reductase (RR, which consists of RRM1 and RRM2 subunits) catalyzes the conversion of ribonucleoside diphosphate into 2'-deoxyribonucleoside diphosphate to maintain the homeostasis of nucleotide pools, which play essential roles in DNA synthesis and DNA repair. In this study, we identified that CDK9 protein expression in adjacent non-tumor tissues predicted HCC patients' overall and progression-free survivals. The anticancer activity of a CDK9-selective inhibitor, LDC000067, on HCC cells was positively associated with its ability to inhibit the expression of RRM1 and RRM2. LDC000067 downregulated RRM1 and RRM2 expression through post-transcriptional pathway. Specifically, LDC000067 triggered RRM2 protein degradation via multiple pathways, including proteasome-, lysosome-, and calcium-dependent pathways. Furthermore, CDK9 positively correlates with RRM1 or RRM2 expression in HCC patients, and the expressions of these three genes were associated with the higher infiltration of immune cells in HCC. Taken together, this study identified the prognostic relevance of CDK9 in HCC and the molecular mechanism for the anticancer effect of CDK9 inhibitors on HCC.

Novel gene signature for predicting biochemical recurrence-free survival of prostate cancer and PRAME modulates prostate cancer progression

Biochemical recurrence (BCR) is considered as an early sign of prostate cancer (PCa) progression after initial treatment, such as radical prostatectomy and radiotherapy; hence, it is important to stratify patients at risk of BCR. In this study, we established a robust 8-gene signature (APOF, Clorf64, RPE65, SEMG1, ARHGDIG, COMP, MKI67 and PRAME) based on the PCa transcriptome profiles in the Cancer Genome Atlas (TCGA) for predicting BCR-free survival of PCa, which was further validated in the MSK-IMPACT Clinical Sequencing Cohort (MSKCC) PCa cohort. Moreover, we found that one risk-related gene (PRAME) was upregulated in tumor samples, particularly in high-risk group was well as in patients metastatic tumor and was correlated with chemotherapeutic drug response. In vitro experiments showed that knocking down PRAME reduced the proliferation, migration, and invasion of PCa cells. Therefore, our study established a new 8-gene signature that could accurately predict the BCR risk of PCa. Inhibition of PRAME attenuated the proliferation, invasion, and migration of PCa cells. These findings provide a novel tool for stratifying high-risk PCa patient and shed light on the mechanism of PCa progression.

Feature selection translates drug response predictors from cell lines to patients

Targeted therapies and chemotherapies are prevalent in cancer treatment. Identification of predictive markers to stratify cancer patients who will respond to these therapies remains challenging because patient drug response data are limited. As large amounts of drug response data have been generated by cell lines, methods to efficiently translate cell-line-trained predictors to human tumors will be useful in clinical practice. Here, we propose versatile feature selection procedures that can be combined with any classifier. For demonstration, we combined the feature selection procedures with a (linear) logit model and a (non-linear) K-nearest neighbor and trained these on cell lines to result in LogitDA and KNNDA, respectively. We show that LogitDA/KNNDA significantly outperforms existing methods, e.g., a logistic model and a deep learning method trained by thousands of genes, in prediction AUC (0.70-1.00 for seven of the ten drugs tested) and is interpretable. This may be due to the fact that sample sizes are often limited in the area of drug response prediction. We further derive a novel adjustment on the prediction cutoff for LogitDA to yield a prediction accuracy of 0.70-0.93 for seven drugs, including erlotinib and cetuximab, whose pathways relevant to anti-cancer therapies are also uncovered. These results indicate that our methods can efficiently translate cell-line-trained predictors into tumors.

iBT-Net: an incremental broad transformer network for cancer drug response prediction

In modern precision medicine, it is an important research topic to predict cancer drug response. Due to incomplete chemical structures and complex gene features, however, it is an ongoing work to design efficient data-driven methods for predicting drug response. Moreover, since the clinical data cannot be easily obtained all at once, the data-driven methods may require relearning when new data are available, resulting in increased time consumption and cost. To address these issues, an incremental broad Transformer network (iBT-Net) is proposed for cancer drug response prediction. Different from the gene expression features learning from cancer cell lines, structural features are further extracted from drugs by Transformer. Broad learning system is then designed to integrate the learned gene features and structural features of drugs to predict the response. With the capability of incremental learning, the proposed method can further use new data to improve its prediction performance without retraining totally. Experiments and comparison studies demonstrate the effectiveness and superiority of iBT-Net under different experimental configurations and continuous data learning.

A landscape of response to drug combinations in non-small cell lung cancer

Combination of anti-cancer drugs is broadly seen as way to overcome the often-limited efficacy of single agents. The design and testing of combinations are however very challenging. Here we present a uniquely large dataset screening over 5000 targeted agent combinations across 81 non-small cell lung cancer cell lines. Our analysis reveals a profound heterogeneity of response across the tumor models. Notably, combinations very rarely result in a strong gain in efficacy over the range of response observable with single agents. Importantly, gain of activity over single agents is more often seen when co-targeting functionally proximal genes, offering a strategy for designing more efficient combinations. Because combinatorial effect is strongly context specific, tumor specificity should be achievable. The resource provided, together with an additional validation screen sheds light on major challenges and opportunities in building efficacious combinations against cancer and provides an opportunity for training computational models for synergy prediction.

PHGDH promotes esophageal squamous cell carcinoma progression via Wnt/β-catenin pathway

PURPOSE

Esophageal squamous carcinoma (ESCC) with a high incidence in China, lacks effective therapeutic targets. Phosphoglycerate dehydrogenase (PHGDH) is a key enzyme in serine biosynthesis. However, the biological role of PHGDH in ESCC has not been revealed.

METHODS

The expression of PHGDH in ESCC was investigated by UALCAN. The relationship between PHGDH expression and its prognostic value was analyzed by Kaplan-Meier and univariate Cox regression. Further, the potential functions of PHGDH involved in ESCC were explored through DAVID database and GSEA software. In addition, the expression of PHGDH was verified in ESCC. Then, the effects of PHGDH knockdown on ESCC were evaluated in vitro and in vivo by cell proliferation, clone formation, cell cycle, apoptosis, tube formation assays and ESCC cells derived xenograft model. In addition, western blotting and immunohistochemistry were used to detect the expression of Wnt/β-catenin pathway which was associated with PHGDH.

RESULTS

Bioinformatics analysis found that PHGDH was highly expressed in ESCC, and meaningfully, patients with high PHGDH expression had a poor prognosis. Moreover, the overexpression of PHGDH was verified in ESCC. Afterwards, PHGDH knockdown inhibited the cell proliferation, induced cell cycle arrest and apoptosis in ESCC cells, and inhibited the angiogenesis of HUVECs induced by ESCC conditioned medium, as well as inhibited the growth of xenograft tumor. Mechanistically, PHGDH knockdown inhibited Wnt/β-catenin signaling pathway in ESCC.

CONCLUSION

High expression of PHGDH predicts a poor prognosis for ESCC. PHGDH knockdown inhibits ESCC progression by suppressing Wnt/β-catenin signaling pathway, indicating that PHGDH might be a potential target for ESCC therapy.

SYK-mediated epithelial cell state is associated with response to c-Met inhibitors in c-Met-overexpressing lung cancer

Genomic MET amplification and exon 14 skipping are currently clinically recognized biomarkers for stratifying subsets of non-small cell lung cancer (NSCLC) patients according to the predicted response to c-Met inhibitors (c-Metis), yet the overall clinical benefit of this strategy is quite limited. Notably, c-Met protein overexpression, which occurs in approximately 20-25% of NSCLC patients, has not yet been clearly defined as a clinically useful biomarker. An optimized strategy for accurately classifying patients with c-Met overexpression for decision-making regarding c-Meti treatment is lacking. Herein, we found that SYK regulates the plasticity of cells in an epithelial state and is associated with their sensitivity to c-Metis both in vitro and in vivo in PDX models with c-Met overexpression regardless of MET gene status. Furthermore, TGF-β1 treatment resulted in SYK transcriptional downregulation, increased Sp1-mediated transcription of FRA1, and restored the mesenchymal state, which conferred resistance to c-Metis. Clinically, a subpopulation of NSCLC patients with c-Met overexpression coupled with SYK overexpression exhibited a high response rate of 73.3% and longer progression-free survival with c-Meti treatment than other patients. SYK negativity coupled with TGF-β1 positivity conferred de novo and acquired resistance. In summary, SYK regulates cell plasticity toward a therapy-sensitive epithelial cell state. Furthermore, our findings showed that SYK overexpression can aid in precisely stratifying NSCLC patients with c-Met overexpression regardless of MET alterations and expand the population predicted to benefit from c-Met-targeted therapy.

Computational drug discovery for castration-resistant prostate cancers through in vitro drug response modeling

Prostate cancer (PC) is the most frequently diagnosed malignancy and a leading cause of cancer deaths in US men. Many PC cases metastasize and develop resistance to systemic hormonal therapy, a stage known as castration-resistant prostate cancer (CRPC). Therefore, there is an urgent need to develop effective therapeutic strategies for CRPC. Traditional drug discovery pipelines require significant time and capital input, which highlights a need for novel methods to evaluate the repositioning potential of existing drugs. Here, we present a computational framework to predict drug sensitivities of clinical CRPC tumors to various existing compounds and identify treatment options with high potential for clinical impact. We applied this method to a CRPC patient cohort and nominated drugs to combat resistance to hormonal therapies including abiraterone and enzalutamide. The utility of this method was demonstrated by nomination of multiple drugs that are currently undergoing clinical trials for CRPC. Additionally, this method identified the tetracycline derivative COL-3, for which we validated higher efficacy in an isogenic cell line model of enzalutamide-resistant vs. enzalutamide-sensitive CRPC. In enzalutamide-resistant CRPC cells, COL-3 displayed higher activity for inhibiting cell growth and migration, and for inducing G1-phase cell cycle arrest and apoptosis. Collectively, these findings demonstrate the utility of a computational framework for independent validation of drugs being tested in CRPC clinical trials, and for nominating drugs with enhanced biological activity in models of enzalutamide-resistant CRPC. The efficiency of this method relative to traditional drug development approaches indicates a high potential for accelerating drug development for CRPC.

Treatment for ovarian clear cell carcinoma with combined inhibition of WEE1 and ATR

BACKGROUND

Standard platinum-based therapy for ovarian cancer is inefficient against ovarian clear cell carcinoma (OCCC). OCCC is a distinct subtype of epithelial ovarian cancer. OCCC constitutes 25% of ovarian cancers in East Asia (Japan, Korea, China, Singapore) and 6-10% in Europe and North America. The cancer is characterized by frequent inactivation of ARID1A and 10% of cases of endometriosis progression to OCCC. The aim of this study was to identify drugs that are either FDA-approved or in clinical trials for the treatment of OCCC.

RESULTS

High throughput screening of 166 compounds that are either FDA-approved, in clinical trials or are in pre-clinical studies identified several cytotoxic compounds against OCCC. ARID1A knockdown cells were more sensitive to inhibitors of either mTOR (PP242), dual mTOR/PI3K (GDC0941), ATR (AZD6738) or MDM2 (RG7388) compared to control cells. Also, compounds targeting BH3 domain (AZD4320) and SRC (AZD0530) displayed preferential cytotoxicity against ARID1A mutant cell lines. In addition, WEE1 inhibitor (AZD1775) showed broad cytotoxicity toward OCCC cell lines, irrespective of ARID1A status.

CONCLUSIONS

In a selection of 166 compounds we showed that inhibitors of ATR and WEE1 were cytotoxic against a panel of OCCC cell lines. These two drugs are already in other clinical trials, making them ideal candidates for treatment of OCCC.

Lights and Shadows on the Cancer Multi-Target Inhibitor Rigosertib (ON-01910.Na)

Rigosertib (ON-01910.Na) is a small-molecule member of the novel synthetic benzyl-styryl-sulfonate family. It is currently in phase III clinical trials for several myelodysplastic syndromes and leukemias and is therefore close to clinical translation. The clinical progress of rigosertib has been hampered by a lack of understanding of its mechanism of action, as it is currently considered a multi-target inhibitor. Rigosertib was first described as an inhibitor of the mitotic master regulator Polo-like kinase 1 (Plk1). However, in recent years, some studies have shown that rigosertib may also interact with the PI3K/Akt pathway, act as a Ras-Raf binding mimetic (altering the Ras signaling pathway), as a microtubule destabilizing agent, or as an activator of a stress-induced phospho-regulatory circuit that ultimately hyperphosphorylates and inactivates Ras signaling effectors. Understanding the mechanism of action of rigosertib has potential clinical implications worth exploring, as it may help to tailor cancer therapies and improve patient outcomes.

Sorting nexin 10 sustains PDGF receptor signaling in glioblastoma stem cells via endosomal protein sorting

Glioblastoma is the most malignant primary brain tumor, the prognosis of which remains dismal even with aggressive surgical, medical, and radiation therapies. Glioblastoma stem cells (GSCs) promote therapeutic resistance and cellular heterogeneity due to their self-renewal properties and capacity for plasticity. To understand the molecular processes essential for maintaining GSCs, we performed an integrative analysis comparing active enhancer landscapes, transcriptional profiles, and functional genomics profiles of GSCs and non-neoplastic neural stem cells (NSCs). We identified sorting nexin 10 (SNX10), an endosomal protein sorting factor, as selectively expressed in GSCs compared with NSCs and essential for GSC survival. Targeting SNX10 impaired GSC viability and proliferation, induced apoptosis, and reduced self-renewal capacity. Mechanistically, GSCs utilized endosomal protein sorting to promote platelet-derived growth factor receptor β (PDGFRβ) proliferative and stem cell signaling pathways through posttranscriptional regulation of the PDGFR tyrosine kinase. Targeting SNX10 expression extended survival of orthotopic xenograft-bearing mice, and high SNX10 expression correlated with poor glioblastoma patient prognosis, suggesting its potential clinical importance. Thus, our study reveals an essential connection between endosomal protein sorting and oncogenic receptor tyrosine kinase signaling and suggests that targeting endosomal sorting may represent a promising therapeutic approach for glioblastoma treatment.

The cell senescence regulator p16 is a promising cancer prognostic and immune check-point inhibitor (ICI) therapy biomarker

Cyclin-dependent kinase inhibitor 2A (CDKN2A) encodes the cell senescence regulator protein p16. The expression of p16 raises in cell senescence and has a nuclear regulation in cell aging. Meanwhile, it's also reported to inhibit the aggression of several cancers. But its clinical application and role in cancer immunotherapy needs further investigation. We collected the transcriptional data of pan-cancer and normal human tissues from The Cancer Genome Atlas and the Genotype-Tissue Expression databases. CBioPortal webtool was employed to mine the genomic alteration status of CDKN2A across cancers. Kaplan-Meier method and univariate Cox regression were performed for prognostic assessments across cancers, respectively. Gene Set Enrichment Analysis is the main method used to search the associated cancer hallmarks associated with CDKN2A. TIMER2.0 was used to analyze the immune cell infiltration relevance with CDKN2A in pan-cancer. The associations between CDKN2A and immunotherapy biomarkers or regulators were performed by spearman correlation analysis. We found CDKN2A is overexpressed in most cancers and exhibits prognosis predictive ability in various cancers. In addition, it is significantly correlated with immune-activated hallmarks, cancer immune cell infiltrations and immunoregulators. The most interesting finding is that CDKN2A can significantly predict anti-PDL1 therapy response. Finally, specific inhibitors which correlated with CDKN2A expression in different cancer types were also screened by using Connectivity Map (CMap) tool. The results revealed that CDKN2A acts as a robust cancer prognostic and immunotherapy biomarker. Its function in the regulation of cancer cell senescence might shape the tumor microenvironment and contribute to its predictive ability of immunotherapy.

Combination of furosemide, gold, and dopamine as a potential therapy for breast cancer

Breast cancer is one of the leading causes of death in women worldwide. Initially, it develops in the epithelium of the ducts or lobules of the breast glandular tissues with limited growth and the potential to metastasize. It is a highly heterogeneous malignancy; however, the common molecular mechanisms could help identify new targeted drugs for treating its subtypes. This study uses computational drug repositioning approaches to explore fresh drug candidates for breast cancer treatment. We also implemented reversal gene expression and gene expression-based signatures to explore novel drug candidates computationally. The drug activity profiles and related gene expression changes were acquired from the DrugBank, PubChem, and LINCS databases, and then in silico drug screening, molecular dynamics (MD) simulation, replica exchange MD simulations, and simulated annealing molecular dynamics (SAMD) simulations were conducted to discover and verify the valid drug candidates. We have found that compounds like furosemide, gold, and dopamine showed significant outcomes. Furthermore, the expression of genes related to breast cancer was observed to be reversed by these shortlisted drugs. Therefore, we postulate that combining furosemide, gold, and dopamine would be a potential combination therapy measurement for breast cancer patients.

Pan-Cancer analysis and experimental validation identify the oncogenic nature of ESPL1: Potential therapeutic target in colorectal cancer

Introduction

Extra spindle pole bodies like 1 (ESPL1) are required to continue the cell cycle, and its primary role is to initiate the final segregation of sister chromatids. Although prior research has revealed a link between ESPL1 and the development of cancer, no systematic pan-cancer analysis has been conducted. Combining multi-omics data with bioinformatics, we have thoroughly described the function of ESPL1 in cancer. In addition, we examined the impact of ESPL1 on the proliferation of numerous cancer cell lines. In addition, the connection between ESPL1 and medication sensitivity was verified using organoids obtained from colorectal cancer patients. All these results confirm the oncogene nature of ESPL1.

Methods

Herein, we downloaded raw data from numerous publicly available databases and then applied R software and online tools to explore the association of ESPL1 expression with prognosis, survival, tumor microenvironment, tumor heterogeneity, and mutational profiles. To validate the oncogene nature of ESPL1, we have performed a knockdown of the target gene in various cancer cell lines to verify the effect of ESPL1 on proliferation and migration. In addition, patients’ derived organoids were used to verify drug sensitivity.

Results

The study found that ESPL1 expression was markedly upregulated in tumorous tissues compared to normal tissues, and high expression of ESPL1 was significantly associated with poor prognosis in a range of cancers. Furthermore, the study revealed that tumors with high ESPL1 expression tended to be more heterogeneous based on various tumor heterogeneity indicators. Enrichment analysis showed that ESPL1 is involved in mediating multiple cancer-related pathways. Notably, the study found that interference with ESPL1 expression significantly inhibited the proliferation of tumor cells. Additionally, the higher the expression of ESPL1 in organoids, the greater the sensitivity to PHA-793887, PAC-1, and AZD7762.

Discussion

Taken together, our study provides evidence that ESPL1 may implicate tumorigenesis and disease progression across multiple cancer types, highlighting its potential utility as both a prognostic indicator and therapeutic target.

Ranking Breast Cancer Drugs and Biomarkers Identification Using Machine Learning and Pharmacogenomics

Breast cancer is one of the major causes of death in women worldwide. It is a diverse illness with substantial intersubject heterogeneity, even among individuals with the same type of tumor, and customized therapy has become increasingly important in this sector. Because of the clinical and physical variability of different kinds of breast cancers, multiple staging and classification systems have been developed. As a result, these tumors exhibit a wide range of gene expression and prognostic indicators. To date, no comprehensive investigation of model training procedures on information from numerous cell line screenings has been conducted together with radiation data. We used human breast cancer cell lines and drug sensitivity information from Cancer Cell Line Encyclopedia (CCLE) and Genomics of Drug Sensitivity in Cancer (GDSC) databases to scan for potential drugs using cell line data. The results are further validated through three machine learning approaches: Elastic Net, LASSO, and Ridge. Next, we selected top-ranked biomarkers based on their role in breast cancer and tested them further for their resistance to radiation using the data from the Cleveland database. We have identified six drugs named Palbociclib, Panobinostat, PD-0325901, PLX4720, Selumetinib, and Tanespimycin that significantly perform on breast cancer cell lines. Also, five biomarkers named TNFSF15, DCAF6, KDM6A, PHETA2, and IFNGR1 are sensitive to all six shortlisted drugs and show sensitivity to the radiations. The proposed biomarkers and drug sensitivity analysis are helpful in translational cancer studies and provide valuable insights for clinical trial design.

Comprehensive analysis of pyroptosis-related gene signatures for glioblastoma immune microenvironment and target therapy

Glioblastoma (GBM) is a malignant brain tumour, but its subtypes (mesenchymal, classical, and proneural) show different prognoses. Pyroptosis is a programmed cell death relating to tumour progression, but its association with GBM is poorly understood. In this work, we collected 73 GBM samples (the Xiangya GBM cohort) and reported that pyroptosis involves tumour-microglia interaction and tumour response to interferon-gamma. GBM samples were grouped into different subtypes, cluster 1 and cluster 2, based on pyroptosis-related genes. Cluster 1 samples manifested a worse prognosis and had a more complicated immune landscape than cluster 2 samples. Single-cell RNA-seq data analysis supported that cluster 1 samples respond to interferon-gamma more actively. Moreover, the machine learning algorithm screened several potential compounds, including nutlin-3, for cluster 1 samples as a novel treatment. In vitro experiments supported that cluster 1 cell line, T98G, is more sensitive to nutlin-3 than cluster 2 cell line, LN229. Nutlin-3 can trigger oxidative stress by increasing DHCR24 expression. Moreover, pyroptosis-resistant genes were upregulated in LN229, which may participate against nutlin-3. Therefore, we hypothesis that GBM may be able to upregulate pyroptosis resistant related genes to against nutlin-3-triggered cell death. In summary, we conclude that pyroptosis highly associates with GBM progression, tumour immune landscape, and tumour response to nutlin-3.

Low level of antioxidant capacity biomarkers but not target overexpression predicts vulnerability to ROS-inducing drugs

Despite a strong rationale for why cancer cells are susceptible to redox-targeting drugs, such drugs often face tumor resistance or dose-limiting toxicity in preclinical and clinical studies. An important reason is the lack of specific biomarkers to better select susceptible cancer entities and stratify patients. Using a large panel of lung cancer cell lines, we identified a set of "antioxidant-capacity" biomarkers (ACB), which were tightly repressed, partly by STAT3 and STAT5A/B in sensitive cells, rendering them susceptible to multiple redox-targeting and ferroptosis-inducing drugs. Contrary to expectation, constitutively low ACB expression was not associated with an increased steady state level of reactive oxygen species (ROS) but a high level of nitric oxide, which is required to sustain high replication rates. Using ACBs, we identified cancer entities with a high percentage of patients with favorable ACB expression pattern, making it likely that more responders to ROS-inducing drugs could be stratified for clinical trials.

Deep learning methods for drug response prediction in cancer: Predominant and emerging trends

Cancer claims millions of lives yearly worldwide. While many therapies have been made available in recent years, by in large cancer remains unsolved. Exploiting computational predictive models to study and treat cancer holds great promise in improving drug development and personalized design of treatment plans, ultimately suppressing tumors, alleviating suffering, and prolonging lives of patients. A wave of recent papers demonstrates promising results in predicting cancer response to drug treatments while utilizing deep learning methods. These papers investigate diverse data representations, neural network architectures, learning methodologies, and evaluations schemes. However, deciphering promising predominant and emerging trends is difficult due to the variety of explored methods and lack of standardized framework for comparing drug response prediction models. To obtain a comprehensive landscape of deep learning methods, we conducted an extensive search and analysis of deep learning models that predict the response to single drug treatments. A total of 61 deep learning-based models have been curated, and summary plots were generated. Based on the analysis, observable patterns and prevalence of methods have been revealed. This review allows to better understand the current state of the field and identify major challenges and promising solution paths.

A systematic assessment of deep learning methods for drug response prediction: from in vitro to clinical applications

Drug response prediction is an important problem in personalized cancer therapy. Among various newly developed models, significant improvement in prediction performance has been reported using deep learning methods. However, systematic comparisons of deep learning methods, especially of the transferability from preclinical models to clinical cohorts, are currently lacking. To provide a more rigorous assessment, the performance of six representative deep learning methods for drug response prediction using nine evaluation metrics, including the overall prediction accuracy, predictability of each drug, potential associated factors and transferability to clinical cohorts, in multiple application scenarios was benchmarked. Most methods show promising prediction within cell line datasets, and TGSA, with its lower time cost and better performance, is recommended. Although the performance metrics decrease when applying models trained on cell lines to patients, a certain amount of power to distinguish clinical response on some drugs can be maintained using CRDNN and TGSA. With these assessments, we provide a guidance for researchers to choose appropriate methods, as well as insights into future directions for the development of more effective methods in clinical scenarios.

A Boolean-based machine learning framework identifies predictive biomarkers of HSP90-targeted therapy response in prostate cancer

Precision medicine has emerged as an important paradigm in oncology, driven by the significant heterogeneity of individual patients' tumour. A key prerequisite for effective implementation of precision oncology is the development of companion biomarkers that can predict response to anti-cancer therapies and guide patient selection for clinical trials and/or treatment. However, reliable predictive biomarkers are currently lacking for many anti-cancer therapies, hampering their clinical application. Here, we developed a novel machine learning-based framework to derive predictive multi-gene biomarker panels and associated expression signatures that accurately predict cancer drug sensitivity. We demonstrated the power of the approach by applying it to identify response biomarker panels for an Hsp90-based therapy in prostate cancer, using proteomic data profiled from prostate cancer patient-derived explants. Our approach employs a rational feature section strategy to maximise model performance, and innovatively utilizes Boolean algebra methods to derive specific expression signatures of the marker proteins. Given suitable data for model training, the approach is also applicable to other cancer drug agents in different tumour settings.

CD73 Inhibits cGAS-STING and Cooperates with CD39 to Promote Pancreatic Cancer

The ectonucleotidases CD39 and CD73 catalyze extracellular ATP to immunosuppressive adenosine, and as such, represent potential cancer targets. We investigated biological impacts of CD39 and CD73 in pancreatic ductal adenocarcinoma (PDAC) by studying clinical samples and experimental mouse tumors. Stromal CD39 and tumoral CD73 expression significantly associated with worse survival in human PDAC samples and abolished the favorable prognostic impact associated with the presence of tumor-infiltrating CD8+ T cells. In mouse transplanted KPC tumors, both CD39 and CD73 on myeloid cells, as well as CD73 on tumor cells, promoted polarization of infiltrating myeloid cells towards an M2-like phenotype, which enhanced tumor growth. CD39 on tumor-specific CD8+ T cells and pancreatic stellate cells also suppressed IFNγ production by T cells. Although therapeutic inhibition of CD39 or CD73 alone significantly delayed tumor growth in vivo, targeting of both ectonucleotidases exhibited markedly superior antitumor activity. CD73 expression on human and mouse PDAC tumor cells also protected against DNA damage induced by gemcitabine and irradiation. Accordingly, large-scale pharmacogenomic analyses of human PDAC cell lines revealed significant associations between CD73 expression and gemcitabine chemoresistance. Strikingly, increased DNA damage in CD73-deficient tumor cells associated with activation of the cGAS-STING pathway. Moreover, cGAS expression in mouse KPC tumor cells was required for antitumor activity of the CD73 inhibitor AB680 in vivo. Our study, thus, illuminates molecular mechanisms whereby CD73 and CD39 seemingly cooperate to promote PDAC progression.

Evolution of precision oncology-guided treatment paradigms

Cancer treatment is gradually evolving from the classical use of nonspecific cytotoxic drugs targeting generic mechanisms of cell growth and proliferation. Instead, new "patient-specific treatment paradigms" that are based on an individual patient's tumor-specific molecular features are emerging, and these include "druggable" genomic alterations such as oncogenic driver mutations, downstream activities of cancer-signaling pathways, and the expression of specific genes involved in tumorigenesis and cancer progression. This evolving landscape of making evidence-based treatment decisions forms the foundation of precision oncology, which aims to deliver "the right drug, to the right patient and at the right time". The long-term vision for this approach is to maximize the treatment efficacy while minimizing exposure to ineffective therapy and reducing co-morbidity-related side effects. Successful clinical translation and implementation of this vision have the potential to revolutionize treatment paradigms from predominantly reactive, to more evidence-based, proactive and predictive care. In this article, we review the past and current approaches in precision oncology, and describe their remarkable power and limitations. We also speculate on the evolution of newly emerging methodologies of the future that can be used to address some of the key challenges associated with the existing paradigms. This article is categorized under: Cancer > Genetics/Genomics/Epigenetics Cancer > Molecular and Cellular Physiology Cancer > Computational Models.

miRNome profiling of lung cancer metastases revealed a key role for miRNA-PD-L1 axis in the modulation of chemotherapy response

Locally advanced non-small cell lung cancer (NSCLC) is frequent at diagnosis and requires multimodal treatment approaches. Neoadjuvant chemotherapy (NACT) followed by surgery is the treatment of choice for operable locally advanced NSCLC (Stage IIIA). However, the majority of patients are NACT-resistant and show persistent lymph nodal metastases (LNmets) and an adverse outcome. Therefore, the identification of mechanisms and biomarkers of NACT resistance is paramount for ameliorating the prognosis of patients with Stage IIIA NSCLC. Here, we investigated the miRNome and transcriptome of chemo-naïve LNmets collected from patients with Stage IIIA NSCLC (N = 64). We found that a microRNA signature accurately predicts NACT response. Mechanistically, we discovered a miR-455-5p/PD-L1 regulatory axis which drives chemotherapy resistance, hallmarks metastases with active IFN-γ response pathway (an inducer of PD-L1 expression), and impacts T cells viability and relative abundances in tumor microenvironment (TME). Our data provide new biomarkers to predict NACT response and add molecular insights relevant for improving the management of patients with locally advanced NSCLC.

Androgen-Independent Prostate Cancer Is Sensitive to CDC42-PAK7 Kinase Inhibition

Prostate cancer is a common form of cancer in men, and androgen-deprivation therapy (ADT) is often used as a first-line treatment. However, some patients develop resistance to ADT, and their disease is called castration-resistant prostate cancer (CRPC). Identifying potential therapeutic targets for this aggressive subtype of prostate cancer is crucial. In this study, we show that statins can selectively inhibit the growth of these CRPC tumors that have lost their androgen receptor (AR) and have overexpressed the RNA-binding protein QKI. We found that the repression of microRNA-200 by QKI overexpression promotes the rise of AR-low mesenchymal-like CRPC cells. Using in silico drug/gene perturbation combined screening, we discovered that QKI-overexpressing cancer cells are selectively vulnerable to CDC42-PAK7 inhibition by statins. We also confirmed that PAK7 overexpression is present in prostate cancer that coexists with hyperlipidemia. Our results demonstrate a previously unseen mechanism of action for statins in these QKI-expressing AR-lost CRPCs. This may explain the clinical benefits of the drug and support the development of a biology-driven drug-repurposing clinical trial. This is an important finding that could help improve treatment options for patients with this aggressive form of prostate cancer.

Involvement of CXCL17 and GPR35 in Gastric Cancer Initiation and Progression

The expression of CXC motif chemokine 17 (CXCL17) and its reported membrane receptor G-protein-coupled receptor 35 (GPR35) in different gastric pathological lesions and their clinical implications are largely unknown. In this study, a total of 860 pathological sections were immune-stained with either anti-CXCL17 or anti-GPR35 antibodies. Their expression was scored within the area of the normal gastric gland of non-atrophic gastritis (NAG-NOR), intestinal metaplasia of atrophic gastritis (AG-IM), IM adjacent to GC (GC-IM), and GC tissue. The clinical significance and potential function of CXCL17 and GPR35 were explored using multiple methods. Our results suggested that CXCL17 expression was gradually upregulated during the pathological progress of gastric diseases (NAG-NOR < AG-IM < GC-IM), but significantly downregulated when GC occurred. GPR35 had a similar expression pattern but its expression in GC remained abundant. High CXCL17 expression in GC was associated with less malignant behavior and was an independent biomarker of favorable prognosis. Overexpressing CXCL17 in HGC27 cells significantly upregulated CCL20 expression. TCGA analysis identified that CXCL17 was negatively correlated with some cancer-promoting pathways and involved in inflammatory activities. CTRP analysis revealed that gastric cell lines expressing less CXCL17 and were more sensitive to the CXCR2 inhibitor SB-225002.

Drug Repurposing Applications to Overcome Male Predominance via Targeting G2/M Checkpoint in Human Esophageal Squamous Cell Carcinoma

Esophageal squamous cell carcinoma (ESCC) is strongly characterized by a male predominance with higher mortality rates and worse responses to treatment in males versus females. Despite the role of sex hormones, other causes that may contribute to sex bias in ESCC remain largely unknown, especially as age increases and the hormone difference begins to diminish between sexes. In this study, we analyzed genomics, transcriptomics, and epigenomics from 663 ESCC patients and found that G2/M checkpoint pathway-related sex bias and age bias were significantly present in multi-omics data. In accordance with gene expression patterns across sexes, ten compounds were identified by applying drug repurposing from three drug sensitivity databases: The Connective Map (CMap), Genomics of Drug Sensitivity in Cancer (GDSC), and The Cancer Therapeutic Response Portal (CTRP). MK1775 and decitabine showed better efficacy in two male ESCC cell lines in vitro and in vivo. The drugs' relevance to the transition between G2 and M was especially evident in male cell lines. In our study, we first validated the sex bias of the G2/M checkpoint pathway in ESCC and then determined that G2/M targets may be included in combination therapy for male patients to improve the efficacy of ESCC treatment.