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

Computer-aided drug discovery strategies for novel therapeutics for prostate cancer leveraging next-generating sequencing data

INTRODUCTION

Prostate cancer (PC) is the most common malignancy and accounts for a significant proportion of cancer deaths among men. Although initial therapy success can often be observed in patients diagnosed with localized PC, many patients eventually develop disease recurrence and metastasis. Without effective treatments, patients with aggressive PC display very poor survival. To curb the current high mortality rate, many investigations have been carried out to identify efficacious therapeutics. Compared to de novo drug designs, computational methods have been widely employed to offer actionable drug predictions in a fast and cost-efficient way. Particularly, powered by an increasing availability of next-generation sequencing molecular profiles from PC patients, computer-aided approaches can be tailored to screen for candidate drugs.

AREAS COVERED

Herein, the authors review the recent advances in computational methods for drug discovery utilizing molecular profiles from PC patients. Given the uniqueness in PC therapeutic needs, they discuss in detail the drug discovery goals of these studies, highlighting their translational values for clinically impactful drug nomination.

EXPERT OPINION

Evolving molecular profiling techniques may enable new perspectives for computer-aided approaches to offer drug candidates for different tumor microenvironments. With ongoing efforts to incorporate new compounds into large-scale high-throughput screens, the authors envision continued expansion of drug candidate pools.

Deep representation learning of chemical-induced transcriptional profile for phenotype-based drug discovery

Artificial intelligence transforms drug discovery, with phenotype-based approaches emerging as a promising alternative to target-based methods, overcoming limitations like lack of well-defined targets. While chemical-induced transcriptional profiles offer a comprehensive view of drug mechanisms, inherent noise often obscures the true signal, hindering their potential for meaningful insights. Here, we highlight the development of TranSiGen, a deep generative model employing self-supervised representation learning. TranSiGen analyzes basal cell gene expression and molecular structures to reconstruct chemical-induced transcriptional profiles with high accuracy. By capturing both cellular and compound information, TranSiGen-derived representations demonstrate efficacy in diverse downstream tasks like ligand-based virtual screening, drug response prediction, and phenotype-based drug repurposing. Notably, in vitro validation of TranSiGen's application in pancreatic cancer drug discovery highlights its potential for identifying effective compounds. We envisage that integrating TranSiGen into the drug discovery and mechanism research holds significant promise for advancing biomedicine.

scRank infers drug-responsive cell types from untreated scRNA-seq data using a target-perturbed gene regulatory network

Cells respond divergently to drugs due to the heterogeneity among cell populations. Thus, it is crucial to identify drug-responsive cell populations in order to accurately elucidate the mechanism of drug action, which is still a great challenge. Here, we address this problem with scRank, which employs a target-perturbed gene regulatory network to rank drug-responsive cell populations via in silico drug perturbations using untreated single-cell transcriptomic data. We benchmark scRank on simulated and real datasets, which shows the superior performance of scRank over existing methods. When applied to medulloblastoma and major depressive disorder datasets, scRank identifies drug-responsive cell types that are consistent with the literature. Moreover, scRank accurately uncovers the macrophage subpopulation responsive to tanshinone IIA and its potential targets in myocardial infarction, with experimental validation. In conclusion, scRank enables the inference of drug-responsive cell types using untreated single-cell data, thus providing insights into the cellular-level impacts of therapeutic interventions.

Integration of Pan-Cancer Cell Line and Single-Cell Transcriptomic Profiles Enables Inference of Therapeutic Vulnerabilities in Heterogeneous Tumors

Single-cell RNA sequencing (scRNA-seq) greatly advanced the understanding of intratumoral heterogeneity by identifying distinct cancer cell subpopulations. However, translating biological differences into treatment strategies is challenging due to a lack of tools to facilitate efficient drug discovery that tackles heterogeneous tumors. Developing such approaches requires accurate prediction of drug response at the single-cell level to offer therapeutic options to specific cell subpopulations. Here, we developed 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 data sets. This method achieved high accuracy in separating cells into their correct cellular drug response statuses. In three distinct prospective tests covering different diseases (rhabdomyosarcoma, pancreatic ductal adenocarcinoma, and castration-resistant prostate cancer), the predicted results using scIDUC were accurate and mirrored biological expectations. In the first two tests, the framework identified drugs for cell subpopulations that were resistant to standard-of-care (SOC) therapies due to intrinsic resistance or tumor microenvironmental effects, and the results showed high consistency with experimental findings from the original studies. In the third test using newly generated SOC therapy-resistant cell lines, scIDUC identified efficacious drugs for the resistant line, and the predictions were validated with in vitro experiments. Together, this study demonstrates the potential of scIDUC to quickly translate scRNA-seq data into drug responses for individual cells, displaying the potential as a tool to improve the treatment of heterogenous tumors.

SIGNIFICANCE

A versatile method that infers cell-level drug response in scRNA-seq data facilitates the development of therapeutic strategies to target heterogeneous subpopulations within a tumor and address issues such as treatment failure and resistance.

Metabolism pathway-based subtyping in endometrial cancer: An integrated study by multi-omics analysis and machine learning algorithms

Endometrial cancer (EC), the second most common malignancy in the female reproductive system, has garnered increasing attention for its genomic heterogeneity, but understanding of its metabolic characteristics is still poor. We explored metabolic dysfunctions in EC through a comprehensive multi-omics analysis (RNA-seq datasets from The Cancer Genome Atlas [TCGA], Cancer Cell Line Encyclopedia [CCLE], and GEO datasets; the Clinical Proteomic Tumor Analysis Consortium [CPTAC] proteomics; CCLE metabolomics) to develop useful molecular targets for precision therapy. Unsupervised consensus clustering was performed to categorize EC patients into three metabolism-pathway-based subgroups (MPSs). These MPS subgroups had distinct clinical prognoses, transcriptomic and genomic alterations, immune microenvironment landscape, and unique patterns of chemotherapy sensitivity. Moreover, the MPS2 subgroup had a better response to immunotherapy. Finally, three machine learning algorithms (LASSO, random forest, and stepwise multivariate Cox regression) were used for developing a prognostic metagene signature based on metabolic molecules. Thus, a 13-hub gene-based classifier was constructed to predict patients' MPS subtypes, offering a more accessible and practical approach. This metabolism-based classification system can enhance prognostic predictions and guide clinical strategies for immunotherapy and metabolism-targeted therapy in EC.

PERCEPTION predicts patient response and resistance to treatment using single-cell transcriptomics of their tumors

Tailoring optimal treatment for individual cancer patients remains a significant challenge. To address this issue, we developed PERCEPTION (PERsonalized Single-Cell Expression-Based Planning for Treatments In ONcology), a precision oncology computational pipeline. Our approach uses publicly available matched bulk and single-cell (sc) expression profiles from large-scale cell-line drug screens. These profiles help build treatment response models based on patients' sc-tumor transcriptomics. PERCEPTION demonstrates success in predicting responses to targeted therapies in cultured and patient-tumor-derived primary cells, as well as in two clinical trials for multiple myeloma and breast cancer. It also captures the resistance development in patients with lung cancer treated with tyrosine kinase inhibitors. PERCEPTION outperforms published state-of-the-art sc-based and bulk-based predictors in all clinical cohorts. PERCEPTION is accessible at https://github.com/ruppinlab/PERCEPTION . Our work, showcasing patient stratification using sc-expression profiles of their tumors, will encourage the adoption of sc-omics profiling in clinical settings, enhancing precision oncology tools based on sc-omics.

Unveiling the hidden role of disulfidptosis in kidney renal clear cell carcinoma: a prognostic signature for personalized treatment

The role of disulfidptosis in kidney renal clear cell carcinoma (KIRC) remains unknown. This study investigated disulfidptosis-related biomarkers for KIRC prognosis prediction and individualized treatment. KIRC patients were clustered by disulfidptosis profiles. Differential expression analysis, survival models, and machine learning were used to construct the disulfidptosis-related prognostic signature (DRPS). Characterizations of the tumor immune microenvironment, genetic drivers, drug sensitivity, and immunotherapy response were explored according to the DRPS risk stratification. Markers included in the signature were validated using single-cell, spatial transcriptomics, quantitative RT-qPCR, and immunohistochemistry. In the discovery cohort, we unveiled two clusters of KIRC patients that differed significantly in disulfidptosis regulator expressions and overall survival (OS). After multiple feature selection steps, a DRPS prognostic model with four features (CHAC1, COL7A1, FOXM1, SHOX2) was constructed and validated. Combined with clinical factors, the model demonstrated robust performance in the discovery and external validation cohorts (5-year AUC = 0.793 and 0.846, respectively). KIRC patients with high-risk scores are characterized by inferior OS, less tumor purity, and increased infiltrations of fibroblasts, M1 macrophages, and B cells. High-risk patients also have higher frequencies of BAP1 and AHNAK2 mutation. Besides, the correlation between the DRPS score and the chemotherapy-response signature indicated the potential effect of Gefitinib for high-risk patients. Among the signature genes, FOXM1 is highly expressed in cycling tumor cells and exhibits spatial aggregation, while others are expressed sparsely within tumor samples. The DRPS model enables improved clinical management and personalized KIRC therapy. The identified biomarkers and immune characteristics offer new mechanistic insight into disulfidptosis in KIRC.

Targeting neddylation and sumoylation in chemoresistant triple negative breast cancer

Triple negative breast cancer (TNBC) accounts for 15-20% of breast cancer cases in the United States. Systemic neoadjuvant chemotherapy (NACT), with or without immunotherapy, is the current standard of care for patients with early-stage TNBC. However, up to 70% of TNBC patients have significant residual disease once NACT is completed, which is associated with a high risk of developing recurrence within two to three years of surgical resection. To identify targetable vulnerabilities in chemoresistant TNBC, we generated longitudinal patient-derived xenograft (PDX) models from TNBC tumors before and after patients received NACT. We then compiled transcriptomes and drug response profiles for all models. Transcriptomic analysis identified the enrichment of aberrant protein homeostasis pathways in models from post-NACT tumors relative to pre-NACT tumors. This observation correlated with increased sensitivity in vitro to inhibitors targeting the proteasome, heat shock proteins, and neddylation pathways. Pevonedistat, a drug annotated as a NEDD8-activating enzyme (NAE) inhibitor, was prioritized for validation in vivo and demonstrated efficacy as a single agent in multiple PDX models of TNBC. Pharmacotranscriptomic analysis identified a pathway-level correlation between pevonedistat activity and post-translational modification (PTM) machinery, particularly involving neddylation and sumoylation targets. Elevated levels of both NEDD8 and SUMO1 were observed in models exhibiting a favorable response to pevonedistat compared to those with a less favorable response in vivo. Moreover, a correlation emerged between the expression of neddylation-regulated pathways and tumor response to pevonedistat, indicating that targeting these PTM pathways may prove effective in combating chemoresistant TNBC.

Improving Anticancer Drug Selection and Prioritization via Neural Learning to Rank

Personalized cancer treatment requires a thorough understanding of complex interactions between drugs and cancer cell lines in varying genetic and molecular contexts. To address this, high-throughput screening has been used to generate large-scale drug response data, facilitating data-driven computational models. Such models can capture complex drug-cell line interactions across various contexts in a fully data-driven manner. However, accurately prioritizing the most effective drugs for each cell line still remains a significant challenge. To address this, we developed multiple neural ranking approaches that leverage large-scale drug response data across multiple cell lines from diverse cancer types. Unlike existing approaches that primarily utilize regression and classification techniques for drug response prediction, we formulated the objective of drug selection and prioritization as a drug ranking problem. In this work, we proposed multiple pairwise and listwise neural ranking methods that learn latent representations of drugs and cell lines and then use those representations to score drugs in each cell line via a learnable scoring function. Specifically, we developed neural pairwise and listwise ranking methods, Pair-PushC and List-One on top of the existing methods, pLETORg and ListNet, respectively. Additionally, we proposed a novel listwise ranking method, List-All, that focuses on all the effective drugs instead of the top effective drug, unlike List-One. We also provide an exhaustive empirical evaluation with state-of-the-art regression and ranking baselines on large-scale data sets across multiple experimental settings. Our results demonstrate that our proposed ranking methods mostly outperform the best baselines with significant improvements of as much as 25.6% in terms of selecting truly effective drugs within the top 20 predicted drugs (i.e., hit@20) across 50% test cell lines. Furthermore, our analyses suggest that the learned latent spaces from our proposed methods demonstrate informative clustering structures and capture relevant underlying biological features. Moreover, our comprehensive evaluation provides a thorough and objective comparison of the performance of different methods (including our proposed ones).

Targeting SOX13 inhibits assembly of respiratory chain supercomplexes to overcome ferroptosis resistance in gastric cancer

Therapeutic resistance represents a bottleneck to treatment in advanced gastric cancer (GC). Ferroptosis is an iron-dependent form of non-apoptotic cell death and is associated with anti-cancer therapeutic efficacy. Further investigations are required to clarify the underlying mechanisms. Ferroptosis-resistant GC cell lines are constructed. Dysregulated mRNAs between ferroptosis-resistant and parental cell lines are identified. The expression of SOX13/SCAF1 is manipulated in GC cell lines where relevant biological and molecular analyses are performed. Molecular docking and computational screening are performed to screen potential inhibitors of SOX13. We show that SOX13 boosts protein remodeling of electron transport chain (ETC) complexes by directly transactivating SCAF1. This leads to increased supercomplexes (SCs) assembly, mitochondrial respiration, mitochondrial energetics and chemo- and immune-resistance. Zanamivir, reverts the ferroptosis-resistant phenotype via directly targeting SOX13 and promoting TRIM25-mediated ubiquitination and degradation of SOX13. Here we show, SOX13/SCAF1 are important in ferroptosis-resistance, and targeting SOX13 with zanamivir has therapeutic potential.

Revealing prognostic insights of programmed cell death (PCD)-associated genes in advanced non-small cell lung cancer

The management of patients with advanced non-small cell lung cancer (NSCLC) presents significant challenges due to cancer cells' intricate and heterogeneous nature. Programmed cell death (PCD) pathways are crucial in diverse biological processes. Nevertheless, the prognostic significance of cell death in NSCLC remains incompletely understood. Our study aims to investigate the prognostic importance of PCD genes and their ability to precisely stratify and evaluate the survival outcomes of patients with advanced NSCLC. We employed Weighted Gene Co-expression Network Analysis (WGCNA), Least Absolute Shrinkage and Selection Operator (LASSO), univariate and multivariate Cox regression analyses for prognostic gene screening. Ultimately, we identified seven PCD-related genes to establish the PCD-related risk score for the advanced NSCLC model (PRAN), effectively stratifying overall survival (OS) in patients with advanced NSCLC. Multivariate Cox regression analysis revealed that the PRAN was the independent prognostic factor than clinical baseline factors. It was positively related to specific metabolic pathways, including hexosamine biosynthesis pathways, which play crucial roles in reprogramming cancer cell metabolism. Furthermore, drug prediction for different PRAN risk groups identified several sensitive drugs explicitly targeting the cell death pathway. Molecular docking analysis suggested the potential therapeutic efficacy of navitoclax in NSCLC, as it demonstrated strong binding with the amino acid residues of C-C motif chemokine ligand 14 (CCL14), carboxypeptidase A3 (CPA3), and C-X3-C motif chemokine receptor 1 (CX3CR1) proteins. The PRAN provides a robust personalized treatment and survival assessment tool in advanced NSCLC patients. Furthermore, identifying sensitive drugs for distinct PRAN risk groups holds promise for advancing targeted therapies in NSCLC.

Identification of crucial ubiquitin-associated genes for predicting the effects of immunotherapy and therapeutic agents in colorectal cancer

BACKGROUND

A growing body of research indicates that colorectal cancer (CRC) is significantly influenced by the ubiquitin-proteasome system. Nevertheless, reliable immune landscapes and ubiquitin-associated prognostic markers are still scarce.

METHODS

We systematically analyzed the RNA-seq data of 2,830 ubiquitin-related genes from Gene Expression Omnibus (GEO) and The Cancer Genome Atlas (TCGA). A CRC prognostic risk model was developed based on ubiquitin-associated gene signatures. In-depth multi-dimensional analyses were performed on ubiquitin-related subgroups with high and low risk. Drug response sensitivity for high-risk CRC patients was also predicted.

RESULTS

A total of 131 ubiquitin-related differentially expressed genes were retrieved, of which 9 prognostic genes for CRC were ultimately identified and further validated by our clinical CRC tumor and adjacent normal samples. The expression pattern of these 9 ubiquitin-associated genes was found to be strongly related to overall survival, immune cell fractions, and immune-related genes of CRC patients. CRC patients stratified by the ubiquitin prognostic model exhibited distinct clinicopathological characteristics and immune landscapes. A comprehensive framework for personalized medicine prediction identified regorafenib and sorafenib as the most promising therapeutic agents for high ubiquitin-related risk CRC patients, which was confirmed in cell viability assays.

CONCLUSIONS

Ubiquitin characteristics can reflect CRC prognosis and help develop innovative biomarkers for precision treatment.

Computational Modeling to Identify Drugs Targeting Metastatic Castration-Resistant Prostate Cancer Characterized by Heightened Glycolysis

Metastatic castration-resistant prostate cancer (mCRPC) remains a deadly disease due to a lack of efficacious treatments. The reprogramming of cancer metabolism toward elevated glycolysis is a hallmark of mCRPC. Our goal is to identify therapeutics specifically associated with high glycolysis. Here, we established a computational framework to identify new pharmacological agents for mCRPC with heightened glycolysis activity under a tumor microenvironment, followed by in vitro validation. First, using our established computational tool, OncoPredict, we imputed the likelihood of drug responses to approximately 1900 agents in each mCRPC tumor from two large clinical patient cohorts. We selected drugs with predicted sensitivity highly correlated with glycolysis scores. In total, 77 drugs predicted to be more sensitive in high glycolysis mCRPC tumors were identified. These drugs represent diverse mechanisms of action. Three of the candidates, ivermectin, CNF2024, and P276-00, were selected for subsequent vitro validation based on the highest measured drug responses associated with glycolysis/OXPHOS in pan-cancer cell lines. By decreasing the input glucose level in culture media to mimic the mCRPC tumor microenvironments, we induced a high-glycolysis condition in PC3 cells and validated the projected higher sensitivity of all three drugs under this condition (p < 0.0001 for all drugs). For biomarker discovery, ivermectin and P276-00 were predicted to be more sensitive to mCRPC tumors with low androgen receptor activities and high glycolysis activities (AR(low)Gly(high)). In addition, we integrated a protein-protein interaction network and topological methods to identify biomarkers for these drug candidates. EEF1B2 and CCNA2 were identified as key biomarkers for ivermectin and CNF2024, respectively, through multiple independent biomarker nomination pipelines. In conclusion, this study offers new efficacious therapeutics beyond traditional androgen-deprivation therapies by precisely targeting mCRPC with high glycolysis.

miR-17-92a-1 cluster host gene: a key regulator in colorectal cancer development and progression

Colorectal cancer (CRC), recognized among the five most prevalent malignancies and most deadly cancers, manifests multifactorial influences stemming from environmental exposures, dietary patterns, age, and genetic predisposition. Although substantial progress has been made in comprehending the etiology of CRC, the precise genetic components driving its pathogenesis remain incompletely elucidated. Within the expansive repertoire of non-coding RNAs, particular focus has centered on the miR-17-92a-1 cluster host gene (MIR17HG) and its associated miRNAs, which actively participate in diverse cellular processes and frequently exhibit heightened expression in various solid tumors, notably CRC. Therefore, the primary objective of this research is to undertake an extensive inquiry into the regulatory mechanisms, structural features, functional attributes, and potential diagnostic and therapeutic implications associated with this cluster in CRC. Furthermore, the intricate interplay between this cluster and the development and progression of CRC will be explored. Our findings underscore the upregulation of the miR-17-92a-1 cluster host gene (MIR17HG) and its associated miRNAs in CRC compared to normal tissues, thus implying their profound involvement in the progression of CRC. Collectively, these molecules are implicated in critical oncogenic processes, encompassing metastatic activity, regulation of apoptotic pathways, cellular proliferation, and drug resistance. Consequently, these findings shed illuminating insights into the potential of MIR17HG and its associated miRNAs as promising targets for therapeutic interventions in the management of CRC.

Overcoming limitations in current measures of drug response may enable AI-driven precision oncology

Machine learning (ML) models of drug sensitivity prediction are becoming increasingly popular in precision oncology. Here, we identify a fundamental limitation in standard measures of drug sensitivity that hinders the development of personalized prediction models - they focus on absolute effects but do not capture relative differences between cancer subtypes. Our work suggests that using z-scored drug response measures mitigates these limitations and leads to meaningful predictions, opening the door for sophisticated ML precision oncology models.

Improving the power of drug toxicity measurements by quantitative nuclei imaging

Imaging-based anticancer drug screens are becoming more prevalent due to development of automated fluorescent microscopes and imaging stations, as well as rapid advancements in image processing software. Automated cell imaging provides many benefits such as their ability to provide high-content data, modularity, dynamics recording and the fact that imaging is the most direct way to access cell viability and cell proliferation. However, currently most publicly available large-scale anticancer drugs screens, such as GDSC, CTRP and NCI-60, provide cell viability data measured by assays based on colorimetric or luminometric measurements of NADH or ATP levels. Although such datasets provide valuable data, it is unclear how well drug toxicity measurements can be integrated with imaging data. Here we explored the relations between drug toxicity data obtained by XTT assay, two quantitative nuclei imaging methods and trypan blue dye exclusion assay using a set of four cancer cell lines with different morphologies and 30 drugs with different mechanisms of action. We show that imaging-based approaches provide high accuracy and the differences between results obtained by different methods highly depend on drug mechanism of action. Selecting AUC metrics over IC50 or comparing data where significantly drugs reduced cell numbers noticeably improves consistency between methods. Using automated cell segmentation protocols we analyzed mitochondria activity in more than 11 thousand drug-treated cells and showed that XTT assay produces unreliable data for CDK4/6, Aurora A, VEGFR and PARP inhibitors due induced cell size growth and increase in individual mitochondria activity. We also explored several benefits of image-based analysis such as ability to monitor cell number dynamics, dissect changes in total and individual mitochondria activity from cell proliferation, and ability to identify chromatin remodeling drugs. Finally, we provide a web tool that allows comparing results obtained by different methods.

Integration of 3D bioprinting and multi-algorithm machine learning identified glioma susceptibilities and microenvironment characteristics

Glioma, with its heterogeneous microenvironments and genetic subtypes, presents substantial challenges for treatment prediction and development. We integrated 3D bioprinting and multi-algorithm machine learning as a novel approach to enhance the assessment and understanding of glioma treatment responses and microenvironment characteristics. The bioprinted patient-derived glioma tissues successfully recapitulated molecular properties and drug responses of native tumors. We then developed GlioML, a machine learning workflow incorporating nine distinct algorithms and a weighted ensemble model that generated robust gene expression-based predictors, each reflecting the diverse action mechanisms of various compounds and drugs. The ensemble model superseded the performance of all individual algorithms across diverse in vitro systems, including sphere cultures, complex 3D bioprinted multicellular models, and 3D patient-derived tissues. By integrating bioprinting, the evaluative scope of the treatment expanded to T cell-related therapy and anti-angiogenesis targeted therapy. We identified promising compounds and drugs for glioma treatment and revealed distinct immunosuppressive or angiogenic myeloid-infiltrated tumor microenvironments. These insights pave the way for enhanced therapeutic development for glioma and potentially for other cancers, highlighting the broad application potential of this integrative and translational approach.

Pan-cancer analysis of Sushi domain-containing protein 4 (SUSD4) and validated in colorectal cancer

Sushi domain-containing protein 4 (SUSD4) is a complement regulatory protein whose primary function is to inhibit the complement system, and it is involved in immune regulation. The role of SUSD4 in cancer progression has largely remained elusive. SUSD4 was studied across a variety of cancer types in this study. According to the results, there is an association between the expression level of SUSD4 and prognosis in multiple types of cancer. Further analysis demonstrated that SUSD4 expression level was related to immune cell infiltration, immune-related genes, tumor heterogeneity, and multiple cancer pathways. Additionally, we validated the function of SUSD4 in colorectal cancer cell lines and found that knockdown of SUSD4 inhibited cell growth and impacted the JAK/STAT pathway. By characterizing drug sensitivity in organoids, we found that the expression of SUSD4 showed a positive correlation trend with IC50 of Selumetinib, YK-4-279, and Piperlongumine. In conclusion, SUSD4 is a valuable prognostic indicator for diverse types of cancer, and it has the potential to be a target for cancer therapy.

Drug repurposing screening and mechanism analysis based on human colorectal cancer organoids

Colorectal cancer (CRC) is a highly heterogeneous cancer and exploring novel therapeutic options is a pressing issue that needs to be addressed. Here, we established human CRC tumor-derived organoids that well represent both morphological and molecular heterogeneities of original tumors. To efficiently identify repurposed drugs for CRC, we developed a robust organoid-based drug screening system. By combining the repurposed drug library and computation-based drug prediction, 335 drugs were tested and 34 drugs with anti-CRC effects were identified. More importantly, we conducted a detailed transcriptome analysis of drug responses and divided the drug response signatures into five representative patterns: differentiation induction, growth inhibition, metabolism inhibition, immune response promotion, and cell cycle inhibition. The anticancer activities of drug candidates were further validated in the established patient-derived organoids-based xenograft (PDOX) system in vivo. We found that fedratinib, trametinib, and bortezomib exhibited effective anticancer effects. Furthermore, the concordance and discordance of drug response signatures between organoids in vitro and pairwise PDOX in vivo were evaluated. Our study offers an innovative approach for drug discovery, and the representative transcriptome features of drug responses provide valuable resources for developing novel clinical treatments for CRC.

An integrated framework for prognosis prediction and drug response modeling in colorectal liver metastasis drug discovery

BACKGROUND

Colorectal cancer (CRC) is the third most prevalent cancer globally, and liver metastasis (CRLM) is the primary cause of death. Hence, it is essential to discover novel prognostic biomarkers and therapeutic drugs for CRLM.

METHODS

This study developed two liver metastasis-associated prognostic signatures based on differentially expressed genes (DEGs) in CRLM. Additionally, we employed an interpretable deep learning model utilizing drug sensitivity databases to identify potential therapeutic drugs for high-risk CRLM patients. Subsequently, in vitro and in vivo experiments were performed to verify the efficacy of these compounds.

RESULTS

These two prognostic models exhibited superior performance compared to previously reported ones. Obatoclax, a BCL-2 inhibitor, showed significant differential responses between high and low risk groups classified by prognostic models, and demonstrated remarkable effectiveness in both Transwell assay and CT26 colorectal liver metastasis mouse model.

CONCLUSIONS

This study highlights the significance of developing specialized prognostication approaches and investigating effective therapeutic drugs for patients with CRLM. The application of a deep learning drug response model provides a new drug discovery strategy for translational medicine in precision oncology.

Computational identification and experimental verification of a novel signature based on SARS-CoV-2-related genes for predicting prognosis, immune microenvironment and therapeutic strategies in lung adenocarcinoma patients

Background

Early research indicates that cancer patients are more vulnerable to adverse outcomes and mortality when infected with severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2). Nonetheless, the specific attributes of SARS-CoV-2 in lung Adenocarcinoma (LUAD) have not been extensively and methodically examined.

Methods

We acquired 322 SARS-CoV-2 infection-related genes (CRGs) from the Human Protein Atlas database. Using an integrative machine learning approach with 10 algorithms, we developed a SARS-CoV-2 score (Cov-2S) signature across The Cancer Genome Atlas and datasets GSE72094, GSE68465, and GSE31210. Comprehensive multi-omics analysis, including assessments of genetic mutations and copy number variations, was conducted to deepen our understanding of the prognosis signature. We also analyzed the response of different Cov-2S subgroups to immunotherapy and identified targeted drugs for these subgroups, advancing personalized medicine strategies. The expression of Cov-2S genes was confirmed through qRT-PCR, with GGH emerging as a critical gene for further functional studies to elucidate its role in LUAD.

Results

Out of 34 differentially expressed CRGs identified, 16 correlated with overall survival. We utilized 10 machine learning algorithms, creating 101 combinations, and selected the RFS as the optimal algorithm for constructing a Cov-2S based on the average C-index across four cohorts. This was achieved after integrating several essential clinicopathological features and 58 established signatures. We observed significant differences in biological functions and immune cell statuses within the tumor microenvironments of high and low Cov-2S groups. Notably, patients with a lower Cov-2S showed enhanced sensitivity to immunotherapy. We also identified five potential drugs targeting Cov-2S. In vitro experiments revealed a significant upregulation of GGH in LUAD, and its knockdown markedly inhibited tumor cell proliferation, migration, and invasion.

Conclusion

Our research has pioneered the development of a consensus Cov-2S signature by employing an innovative approach with 10 machine learning algorithms for LUAD. Cov-2S reliably forecasts the prognosis, mirrors the tumor's local immune condition, and supports clinical decision-making in tumor therapies.

Threonine fuels glioblastoma through YRDC-mediated codon-biased translational reprogramming

Cancers commonly reprogram translation and metabolism, but little is known about how these two features coordinate in cancer stem cells. Here we show that glioblastoma stem cells (GSCs) display elevated protein translation. To dissect underlying mechanisms, we performed a CRISPR screen and identified YRDC as the top essential transfer RNA (tRNA) modification enzyme in GSCs. YRDC catalyzes the formation of N6-threonylcarbamoyladenosine (t6A) on ANN-decoding tRNA species (A denotes adenosine, and N denotes any nucleotide). Targeting YRDC reduced t6A formation, suppressed global translation and inhibited tumor growth both in vitro and in vivo. Threonine is an essential substrate of YRDC. Threonine accumulated in GSCs, which facilitated t6A formation through YRDC and shifted the proteome to support mitosis-related genes with ANN codon bias. Dietary threonine restriction (TR) reduced tumor t6A formation, slowed xenograft growth and augmented anti-tumor efficacy of chemotherapy and anti-mitotic therapy, providing a molecular basis for a dietary intervention in cancer treatment.

Multi-dimensional characterization of apoptosis in the tumor microenvironment and therapeutic relevance in melanoma

PURPOSE

Melanoma is widely utilized as a prominent model for the development of immunotherapy, thought an inadequate immune response can occur. Moreover, the development of apoptosis-related therapies and combinations with other therapeutic strategies is impeded by the limited understanding of apoptosis's role within diverse tumor immune microenvironments (TMEs).

METHODS

Here, we constructed an apoptosis-related tumor microenvironment signature (ATM) and employ multi-dimensional analysis to understand the roles of apoptosis in tumor microenvironment. We further assessed the clinical applications of ATM in nine independent cohorts, and anticipated the impact of ATM on cellular drug response in cultured cells.

RESULTS

Our ATM model exhibits robust performance in survival prediction in multiple melanoma cohorts. Different ATM groups exhibited distinct molecular signatures and biological processes. The low ATM group exhibited significant enrichment in B cell activation-related pathways. What's more, plasma cells showed the lowest ATM score, highlighting their role as pivotal contributors in the ATM model. Mechanistically, the analysis of the interplay between plasma cells and other immune cells elucidated their crucial role in orchestrating an effective anti-tumor immune response. Significantly, the ATM signature exhibited associations with therapeutic efficacy of immune checkpoint blockade and the drug sensitivity of various agents, including FDA-approved and clinically utilized drugs targeting the VEGF signaling pathway. Finally, ATM was associated with tertiary lymphoid structures (TLS), exhibiting stronger patient stratification ability compared to classical "hot tumors".

CONCLUSION

Our findings indicate that ATM is a prognostic factor and is associated with the immune response and drug sensitivity in melanoma.

A deep learning model of tumor cell architecture elucidates response and resistance to CDK4/6 inhibitors

Cyclin-dependent kinase 4 and 6 inhibitors (CDK4/6is) have revolutionized breast cancer therapy. However, <50% of patients have an objective response, and nearly all patients develop resistance during therapy. To elucidate the underlying mechanisms, we constructed an interpretable deep learning model of the response to palbociclib, a CDK4/6i, based on a reference map of multiprotein assemblies in cancer. The model identifies eight core assemblies that integrate rare and common alterations across 90 genes to stratify palbociclib-sensitive versus palbociclib-resistant cell lines. Predictions translate to patients and patient-derived xenografts, whereas single-gene biomarkers do not. Most predictive assemblies can be shown by CRISPR-Cas9 genetic disruption to regulate the CDK4/6i response. Validated assemblies relate to cell-cycle control, growth factor signaling and a histone regulatory complex that we show promotes S-phase entry through the activation of the histone modifiers KAT6A and TBL1XR1 and the transcription factor RUNX1. This study enables an integrated assessment of how a tumor's genetic profile modulates CDK4/6i resistance.

Comprehensive Analyses of PANoptosome with Potential Implications in Cancer Prognosis and Immunotherapy

Cell death resistance significantly contributes to poor therapeutic outcomes in various cancers. PANoptosis, a unique inflammatory programmed cell death (PCD) pathway activated by specific triggers and regulated by the PANoptosome, possesses key features of apoptosis, pyroptosis, and necroptosis, but these cannot be accounted for by any of the three PCD pathways alone. While existing studies on PANoptosis have predominantly centered on infectious and inflammatory diseases, its role in cancer malignancy has been understudied. In this comprehensive investigation, we conducted pan-cancer analyses of PANoptosome component genes across 33 cancer types. We characterized the genetic, epigenetic, and transcriptomic landscapes, and introduced a PANoptosome-related potential index (PANo-RPI) for evaluating the intrinsic PANoptosome assembly potential in cancers. Our findings unveil PANo-RPI as a prognostic factor in numerous cancers, including KIRC, LGG, and PAAD. Crucially, we established a significant correlation between PANo-RPI and tumor immune responses, as well as the infiltration of diverse lymphoid and myeloid cell subsets across nearly all cancer types. Moreover, a high PANo-RPI was consistently associated with improved immunotherapy response and efficacy, as evidenced by re-analysis of multiple immunotherapy cohorts. In conclusion, our study suggests that targeting PANoptosome components and modulating PANoptosis may hold tremendous therapeutic potential in the context of cancer.

Cancer Mutations Converge on a Collection of Protein Assemblies to Predict Resistance to Replication Stress

Rapid proliferation is a hallmark of cancer associated with sensitivity to therapeutics that cause DNA replication stress (RS). Many tumors exhibit drug resistance, however, via molecular pathways that are incompletely understood. Here, we develop an ensemble of predictive models that elucidate how cancer mutations impact the response to common RS-inducing (RSi) agents. The models implement recent advances in deep learning to facilitate multidrug prediction and mechanistic interpretation. Initial studies in tumor cells identify 41 molecular assemblies that integrate alterations in hundreds of genes for accurate drug response prediction. These cover roles in transcription, repair, cell-cycle checkpoints, and growth signaling, of which 30 are shown by loss-of-function genetic screens to regulate drug sensitivity or replication restart. The model translates to cisplatin-treated cervical cancer patients, highlighting an RTK-JAK-STAT assembly governing resistance. This study defines a compendium of mechanisms by which mutations affect therapeutic responses, with implications for precision medicine.

SIGNIFICANCE

Zhao and colleagues use recent advances in machine learning to study the effects of tumor mutations on the response to common therapeutics that cause RS. The resulting predictive models integrate numerous genetic alterations distributed across a constellation of molecular assemblies, facilitating a quantitative and interpretable assessment of drug response. This article is featured in Selected Articles from This Issue, p. 384.

Expression of the checkpoint kinase BUB1 is a predictor of response to cancer therapies

The identification of clinically-relevant biomarkers is of upmost importance for the management of cancer, from diagnosis to treatment choices. We performed a pan-cancer analysis of the mitotic checkpoint budding uninhibited by benzimidazole 1 gene BUB1, in the attempt to ascertain its diagnostic and prognostic values, specifically in the context of drug response. BUB1 was found to be overexpressed in the majority of cancers, and particularly elevated in clinically aggressive molecular subtypes. Its expression was correlated with clinico-phenotypic features, notably tumour staging, size, invasion, hypoxia, and stemness. In terms of prognostic value, the expression of BUB1 bore differential clinical outcomes depending on the treatment administered in TCGA cancer cohorts, suggesting sensitivity or resistance, depending on the expression levels. We also integrated in vitro drug sensitivity data from public projects based on correlation between drug efficacy and BUB1 expression to produce a list of candidate compounds with differential responses according to BUB1 levels. Gene Ontology enrichment analyses revealed that BUB1 overexpression in cancer is associated with biological processes related to mitosis and chromosome segregation machinery, reflecting the mechanisms of action of drugs with a differential effect based on BUB1 expression.

Anthracyclines induce cardiotoxicity through a shared gene expression response signature

TOP2 inhibitors (TOP2i) are effective drugs for breast cancer treatment. However, they can cause cardiotoxicity in some women. The most widely used TOP2i include anthracyclines (AC) Doxorubicin (DOX), Daunorubicin (DNR), Epirubicin (EPI), and the anthraquinone Mitoxantrone (MTX). It is unclear whether women would experience the same adverse effects from all drugs in this class, or if specific drugs would be preferable for certain individuals based on their cardiotoxicity risk profile. To investigate this, we studied the effects of treatment of DOX, DNR, EPI, MTX, and an unrelated monoclonal antibody Trastuzumab (TRZ) on iPSC-derived cardiomyocytes (iPSC-CMs) from six healthy females. All TOP2i induce cell death at concentrations observed in cancer patient serum, while TRZ does not. A sub-lethal dose of all TOP2i induces limited cellular stress but affects calcium handling, a function critical for cardiomyocyte contraction. TOP2i induce thousands of gene expression changes over time, giving rise to four distinct gene expression response signatures, denoted as TOP2i early-acute, early-sustained, and late response genes, and non-response genes. There is no drug- or AC-specific signature. TOP2i early response genes are enriched in chromatin regulators, which mediate AC sensitivity across breast cancer patients. However, there is increased transcriptional variability between individuals following AC treatments. To investigate potential genetic effects on response variability, we first identified a reported set of expression quantitative trait loci (eQTLs) uncovered following DOX treatment in iPSC-CMs. Indeed, DOX response eQTLs are enriched in genes that respond to all TOP2i. Next, we identified 38 genes in loci associated with AC toxicity by GWAS or TWAS. Two thirds of the genes that respond to at least one TOP2i, respond to all ACs with the same direction of effect. Our data demonstrate that TOP2i induce thousands of shared gene expression changes in cardiomyocytes, including genes near SNPs associated with inter-individual variation in response to DOX treatment and AC-induced cardiotoxicity.

A review of computational methods for predicting cancer drug response at the single-cell level through integration with bulk RNAseq data

Cancer treatment failure is often attributed to tumor heterogeneity, where diverse malignant cell clones exist within a patient. Despite a growing understanding of heterogeneous tumor cells depicted by single-cell RNA sequencing (scRNA-seq), there is still a gap in the translation of such knowledge into treatment strategies tackling the pervasive issue of therapy resistance. In this review, we survey methods leveraging large-scale drug screens to generate cellular sensitivities to various therapeutics. These methods enable efficient drug screens in scRNA-seq data and serve as the bedrock of drug discovery for specific cancer cell groups. We envision that they will become an indispensable tool for tailoring patient care in the era of heterogeneity-aware precision medicine.

Repositioning VU-0365114 as a novel microtubule-destabilizing agent for treating cancer and overcoming drug resistance

Microtubule-targeting agents represent one of the most successful classes of anticancer agents. However, the development of drug resistance and the appearance of adverse effects hamper their clinical implementation. Novel microtubule-targeting agents without such limitations are urgently needed. By employing a gene expression-based drug repositioning strategy, this study identifies VU-0365114, originally synthesized as a positive allosteric modulator of human muscarinic acetylcholine receptor M5 (M5 mAChR), as a novel type of tubulin inhibitor by destabilizing microtubules. VU-0365114 exhibits a broad-spectrum in vitro anticancer activity, especially in colorectal cancer cells. A tumor xenograft study in nude mice shows that VU-0365114 slowed the in vivo colorectal tumor growth. The anticancer activity of VU-0365114 is not related to its original target, M5 mAChR. In addition, VU-0365114 does not serve as a substrate of multidrug resistance (MDR) proteins, and thus, it can overcome MDR. Furthermore, a kinome analysis shows that VU-0365114 did not exhibit other significant off-target effects. Taken together, our study suggests that VU-0365114 primarily targets microtubules, offering potential for repurposing in cancer treatment, although more studies are needed before further drug development.

Transcriptional analysis of landmark molecular pathways in lung adenocarcinoma results in a clinically relevant classification with potential therapeutic implications

Lung adenocarcinoma (LUAD) is a molecularly heterogeneous disease. In addition to genomic alterations, cancer transcriptional profiling can be helpful to tailor cancer treatment and to estimate each patient's outcome. Transcriptional activity levels of 50 molecular pathways were inferred in 4573 LUAD patients using Gene Set Variation Analysis (GSVA) method. Seven LUAD subtypes were defined and independently validated based on the combined behavior of the studied pathways: AD (adenocarcinoma subtype) 1-7. AD1, AD4, and AD5 subtypes were associated with better overall survival. AD1 and AD4 subtypes were enriched in epidermal growth factor receptor (EGFR) mutations, whereas AD2 and AD6 showed higher tumor protein p53 (TP53) alteration frequencies. AD2 and AD6 subtypes correlated with higher genome instability, proliferation-related pathway expression, and specific sensitivity to chemotherapy, based on data from LUAD cell lines. LUAD subtypes were able to predict immunotherapy response in addition to CD274 (PD-L1) gene expression and tumor mutational burden (TMB). AD2 and AD4 subtypes were associated with potential resistance and response to immunotherapy, respectively. Thus, analysis of transcriptomic data could improve patient stratification beyond genomics and single biomarkers (i.e., PD-L1 and TMB) and may lay the foundation for more personalized treatment avenues, especially in driver-negative LUAD.

Improving anti-cancer drug response prediction using multi-task learning on graph convolutional networks

Predicting the therapeutic effect of anti-cancer drugs on tumors based on the characteristics of tumors and patients is one of the important contents of precision oncology. Existing computational methods regard the drug response prediction problem as a classification or regression task. However, few of them consider leveraging the relationship between the two tasks. In this work, we propose a Multi-task Interaction Graph Convolutional Network (MTIGCN) for anti-cancer drug response prediction. MTIGCN first utilizes an graph convolutional network-based model to produce embeddings for both cell lines and drugs. After that, the model employs multi-task learning to predict anti-cancer drug response, which involves training the model on three different tasks simultaneously: the main task of the drug sensitive or resistant classification task and the two auxiliary tasks of regression prediction and similarity network reconstruction. By sharing parameters and optimizing the losses of different tasks simultaneously, MTIGCN enhances the feature representation and reduces overfitting. The results of the experiments on two in vitro datasets demonstrated that MTIGCN outperformed seven state-of-the-art baseline methods. Moreover, the well-trained model on the in vitro dataset GDSC exhibited good performance when applied to predict drug responses in in vivo datasets PDX and TCGA. The case study confirmed the model's ability to discover unknown drug responses in cell lines.

Selenium reduction of ubiquinone via SQOR suppresses ferroptosis

The canonical biological function of selenium is in the production of selenocysteine residues of selenoproteins, and this forms the basis for its role as an essential antioxidant and cytoprotective micronutrient. Here we demonstrate that, via its metabolic intermediate hydrogen selenide, selenium reduces ubiquinone in the mitochondria through catalysis by sulfide quinone oxidoreductase. Through this mechanism, selenium rapidly protects against lipid peroxidation and ferroptosis in a timescale that precedes selenoprotein production, doing so even when selenoprotein production has been eliminated. Our findings identify a regulatory mechanism against ferroptosis that implicates sulfide quinone oxidoreductase and expands our understanding of selenium in biology.

Integrated drug response prediction models pinpoint repurposed drugs with effectiveness against rhabdomyosarcoma

Targeted therapies for inhibiting the growth of cancer cells or inducing apoptosis are urgently needed for effective rhabdomyosarcoma (RMS) treatment. However, identifying cancer-targeting compounds with few side effects, among the many potential compounds, is expensive and time-consuming. A computational approach to reduce the number of potential candidate drugs can facilitate the discovery of attractive lead compounds. To address this and obtain reliable predictions of novel cell-line-specific drugs, we apply prediction models that have the potential to improve drug discovery approaches for RMS treatment. The results of two prediction models were ensemble and validated via in vitro experiments. The computational models were trained using data extracted from the Genomics of Drug Sensitivity in Cancer database and tested on two RMS cell lines to select potential RMS drug candidates. Among 235 candidate drugs, 22 were selected following the result of the computational approach, and three candidate drugs were identified (NSC207895, vorinostat, and belinostat) that showed selective effectiveness in RMS cell lines in vitro via the induction of apoptosis. Our in vitro experiments have demonstrated that our proposed methods can effectively identify and repurpose drugs for treating RMS.

PANCDR: precise medicine prediction using an adversarial network for cancer drug response

Pharmacogenomics aims to provide personalized therapy to patients based on their genetic variability. However, accurate prediction of cancer drug response (CDR) is challenging due to genetic heterogeneity. Since clinical data are limited, most studies predicting drug response use preclinical data to train models. However, such models might not be generalizable to external clinical data due to differences between the preclinical and clinical datasets. In this study, a Precision Medicine Prediction using an Adversarial Network for Cancer Drug Response (PANCDR) model is proposed. PANCDR consists of two sub-models, an adversarial model and a CDR prediction model. The adversarial model reduces the gap between the preclinical and clinical datasets, while the CDR prediction model extracts features and predicts responses. PANCDR was trained using both preclinical data and unlabeled clinical data. Subsequently, it was tested on external clinical data, including The Cancer Genome Atlas and brain tumor patients. PANCDR outperformed other machine learning models in predicting external test data. Our results demonstrate the robustness of PANCDR and its potential in precision medicine by recommending patient-specific drug candidates. The PANCDR codes and data are available at https://github.com/DMCB-GIST/PANCDR.

Integrative analysis of co-expression pattern of solute carrier transporters reveals molecular subtypes associated with tumor microenvironment hallmarks and clinical outcomes in colon cancer

Recent findings have suggested that solute carrier (SLC) transporters play an important role in tumor development and progression, and alterations in the expression of individual SLC genes are critical for fulfilling the heightened metabolic requirements of cancerous cells. However, the global influence of the co-expression pattern of SLC transporters on the clinical stratification and characteristics of the tumor microenvironment (TME) remains unexplored. In this study, we identified five SLC gene subtypes based on transcriptome co-expression patterns of 187 SLC transporters by consensus clustering analysis. These subtypes, which were characterized by distinct TME and biological characteristics, were successfully employed for prognostic and chemotherapy response prediction in colon cancer patients, as well as demonstrated associations with immunotherapy benefits. Then, we generated an SLC score model comprising 113 genes to quantify SLC gene co-expression patterns and validated it as an independent prognostic factor and drug response predictor in several independent colon cancer cohorts. Patients with a high SLC score possessed distinct characteristics of copy number variation, genomic mutations, DNA methylation, and indicated an SLC-S2 subtype, which was characterized by strong stromal cell infiltration, stromal pathway activation, poor prognosis, and low predicted fluorouracil and immunotherapeutic responses. Furthermore, the analysis of the Cancer Therapeutics Response Portal database revealed that inhibitors targeting PI3K catalytic subunits could serve as promising chemosensitizing agents for individuals exhibiting high SLC scores. In conclusion, the co-expression patterns of SLC transporters aided the disease classification, and the SLC score proved to be a reliable tool for distinguishing SLC gene subtypes and guiding precise treatment in patients with colon cancer.

Cancer drug sensitivity prediction from routine histology images

Drug sensitivity prediction models can aid in personalising cancer therapy, biomarker discovery, and drug design. Such models require survival data from randomised controlled trials which can be time consuming and expensive. In this proof-of-concept study, we demonstrate for the first time that deep learning can link histological patterns in whole slide images (WSIs) of Haematoxylin & Eosin (H&E) stained breast cancer sections with drug sensitivities inferred from cell lines. We employ patient-wise drug sensitivities imputed from gene expression-based mapping of drug effects on cancer cell lines to train a deep learning model that predicts patients' sensitivity to multiple drugs from WSIs. We show that it is possible to use routine WSIs to predict the drug sensitivity profile of a cancer patient for a number of approved and experimental drugs. We also show that the proposed approach can identify cellular and histological patterns associated with drug sensitivity profiles of cancer patients.

Inhibition of iRhom1 by CD44-targeting nanocarrier for improved cancer immunochemotherapy

The multifaceted chemo-immune resistance is the principal barrier to achieving cure in cancer patients. Identifying a target that is critically involved in chemo-immune-resistance represents an attractive strategy to improve cancer treatment. iRhom1 plays a role in cancer cell proliferation and its expression is negatively correlated with immune cell infiltration. Here we show that iRhom1 decreases chemotherapy sensitivity by regulating the MAPK14-HSP27 axis. In addition, iRhom1 inhibits the cytotoxic T-cell response by reducing the stability of ERAP1 protein and the ERAP1-mediated antigen processing and presentation. To facilitate the therapeutic translation of these findings, we develop a biodegradable nanocarrier that is effective in codelivery of iRhom pre-siRNA (pre-siiRhom) and chemotherapeutic drugs. This nanocarrier is effective in tumor targeting and penetration through both enhanced permeability and retention effect and CD44-mediated transcytosis in tumor endothelial cells as well as tumor cells. Inhibition of iRhom1 further facilitates tumor targeting and uptake through inhibition of CD44 cleavage. Co-delivery of pre-siiRhom and a chemotherapy agent leads to enhanced antitumor efficacy and activated tumor immune microenvironment in multiple cancer models in female mice. Targeting iRhom1 together with chemotherapy could represent a strategy to overcome chemo-immune resistance in cancer treatment.

Illuminating phenotypic drug responses of sarcoma cells to kinase inhibitors by phosphoproteomics

Kinase inhibitors (KIs) are important cancer drugs but often feature polypharmacology that is molecularly not understood. This disconnect is particularly apparent in cancer entities such as sarcomas for which the oncogenic drivers are often not clear. To investigate more systematically how the cellular proteotypes of sarcoma cells shape their response to molecularly targeted drugs, we profiled the proteomes and phosphoproteomes of 17 sarcoma cell lines and screened the same against 150 cancer drugs. The resulting 2550 phenotypic profiles revealed distinct drug responses and the cellular activity landscapes derived from deep (phospho)proteomes (9-10,000 proteins and 10-27,000 phosphorylation sites per cell line) enabled several lines of analysis. For instance, connecting the (phospho)proteomic data with drug responses revealed known and novel mechanisms of action (MoAs) of KIs and identified markers of drug sensitivity or resistance. All data is publicly accessible via an interactive web application that enables exploration of this rich molecular resource for a better understanding of active signalling pathways in sarcoma cells, identifying treatment response predictors and revealing novel MoA of clinical KIs.

Profiling Numerical and Structural Chromosomal Instability in Different Cancer Types

Many cancers display whole chromosome instability (W-CIN) and structural chromosomal instability (S-CIN), referring to increased rates of acquiring numerically and structurally abnormal chromosome changes. This protocol provides detailed steps to analyze the W-CIN and S-CIN across cancer types, intending to leverage large-scale bulk sequencing and SNP array data complemented with the computational models to gain a better understanding of W-CIN and S-CIN.

Comparison of multiple modalities for drug response prediction with learning curves using neural networks and XGBoost

Motivation

Anti-cancer drug response prediction is a central problem within stratified medicine. Transcriptomic profiles of cancer cell lines are typically used for drug response prediction, but we hypothesize that proteomics or phosphoproteomics might be more suitable as they give a more direct insight into cellular processes. However, there has not yet been a systematic comparison between all three of these datatypes using consistent evaluation criteria.

Results

Due to the limited number of cell lines with phosphoproteomics profiles we use learning curves, a plot of predictive performance as a function of dataset size, to compare the current performance and predict the future performance of the three omics datasets with more data. We use neural networks and XGBoost and compare them against a simple rule-based benchmark. We show that phosphoproteomics slightly outperforms RNA-seq and proteomics using the 38 cell lines with profiles of all three omics data types. Furthermore, using the 877 cell lines with proteomics and RNA-seq profiles, we show that RNA-seq slightly outperforms proteomics. With the learning curves we predict that the mean squared error using the phosphoproteomics dataset would decrease by ∼ 15 % if a dataset of the same size as the proteomics/transcriptomics was collected. For the cell lines with proteomics and RNA-seq profiles the learning curves reveal that for smaller dataset sizes neural networks outperform XGBoost and vice versa for larger datasets. Furthermore, the trajectory of the XGBoost curve suggests that it will improve faster than the neural networks as more data are collected.

Availability and implementation

See https://github.com/Nik-BB/Learning-curves-for-DRP for the code used.

TNFR1 signaling converging on FGF14 controls neuronal hyperactivity and sickness behavior in experimental cerebral malaria

BACKGROUND

Excess tumor necrosis factor (TNF) is implicated in the pathogenesis of hyperinflammatory experimental cerebral malaria (eCM), including gliosis, increased levels of fibrin(ogen) in the brain, behavioral changes, and mortality. However, the role of TNF in eCM within the brain parenchyma, particularly directly on neurons, remains underdefined. Here, we investigate electrophysiological consequences of eCM on neuronal excitability and cell signaling mechanisms that contribute to observed phenotypes.

METHODS

The split-luciferase complementation assay (LCA) was used to investigate cell signaling mechanisms downstream of tumor necrosis factor receptor 1 (TNFR1) that could contribute to changes in neuronal excitability in eCM. Whole-cell patch-clamp electrophysiology was performed in brain slices from eCM mice to elucidate consequences of infection on CA1 pyramidal neuron excitability and cell signaling mechanisms that contribute to observed phenotypes. Involvement of identified signaling molecules in mediating behavioral changes and sickness behavior observed in eCM were investigated in vivo using genetic silencing.

RESULTS

Exploring signaling mechanisms that underlie TNF-induced effects on neuronal excitability, we found that the complex assembly of fibroblast growth factor 14 (FGF14) and the voltage-gated Na+ (Nav) channel 1.6 (Nav1.6) is increased upon tumor necrosis factor receptor 1 (TNFR1) stimulation via Janus Kinase 2 (JAK2). On account of the dependency of hyperinflammatory experimental cerebral malaria (eCM) on TNF, we performed patch-clamp studies in slices from eCM mice and showed that Plasmodium chabaudi infection augments Nav1.6 channel conductance of CA1 pyramidal neurons through the TNFR1-JAK2-FGF14-Nav1.6 signaling network, which leads to hyperexcitability. Hyperexcitability of CA1 pyramidal neurons caused by infection was mitigated via an anti-TNF antibody and genetic silencing of FGF14 in CA1. Furthermore, knockdown of FGF14 in CA1 reduced sickness behavior caused by infection.

CONCLUSIONS

FGF14 may represent a therapeutic target for mitigating consequences of TNF-mediated neuroinflammation.

GPDRP: a multimodal framework for drug response prediction with graph transformer

BACKGROUND

In the field of computational personalized medicine, drug response prediction (DRP) is a critical issue. However, existing studies often characterize drugs as strings, a representation that does not align with the natural description of molecules. Additionally, they ignore gene pathway-specific combinatorial implication.

RESULTS

In this study, we propose drug Graph and gene Pathway based Drug response prediction method (GPDRP), a new multimodal deep learning model for predicting drug responses based on drug molecular graphs and gene pathway activity. In GPDRP, drugs are represented by molecular graphs, while cell lines are described by gene pathway activity scores. The model separately learns these two types of data using Graph Neural Networks (GNN) with Graph Transformers and deep neural networks. Predictions are subsequently made through fully connected layers.

CONCLUSIONS

Our results indicate that Graph Transformer-based model delivers superior performance. We apply GPDRP on hundreds of cancer cell lines' bulk RNA-sequencing data, and it outperforms some recently published models. Furthermore, the generalizability and applicability of GPDRP are demonstrated through its predictions on unknown drug-cell line pairs and xenografts. This underscores the interpretability achieved by incorporating gene pathways.

Deep learning untangles the resistance mechanism of p53 reactivator in lung cancer cells

Tumor suppressor p53 plays a pivotal role in suppressing cancer, so various drugs has been suggested to upregulate its function. However, drug resistance is still the biggest hurdle to be overcome. To address this, we developed a deep learning model called AnoDAN (anomalous gene detection using generative adversarial networks and graph neural networks for overcoming drug resistance) that unravels the hidden resistance mechanisms and identifies a combinatorial target to overcome the resistance. Our findings reveal that the TGF-β signaling pathway, alongside the p53 signaling pathway, mediates the resistance, with THBS1 serving as a core regulatory target in both pathways. Experimental validation in lung cancer cells confirms the effects of THBS1 on responsiveness to a p53 reactivator. We further discovered the positive feedback loop between THBS1 and the TGF-β pathway as the main source of resistance. This study enhances our understanding of p53 regulation and offers insights into overcoming drug resistance.

A Web Application for Predicting Drug Combination Efficacy Using Monotherapy Data and IDACombo

We recently reported a computational method (IDACombo) designed to predict the efficacy of cancer drug combinations using monotherapy response data and the assumptions of independent drug action. Given the strong agreement between IDACombo predictions and measured drug combination efficacy in vitro and in clinical trials, we believe IDACombo can be of immediate use to researchers who are working to develop novel drug combinations. While we previously released our method as an R package, we have now created an R Shiny application to allow researchers without programming experience to easily utilize this method. The app provides a graphical interface which enables users to easily generate efficacy predictions with IDACombo using provided data from several high-throughput cell line screens or using custom, user-provided data.

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 informatics 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/.

Large-Scale Information Retrieval and Correction of Noisy Pharmacogenomic Datasets through Residual Thresholded Deep Matrix Factorization

Pharmacogenomics studies are attracting an increasing amount of interest from researchers in precision medicine. The advances in high-throughput experiments and multiplexed approaches allow the large-scale quantification of drug sensitivities in molecularly characterized cancer cell lines (CCLs), resulting in a number of open drug sensitivity datasets for drug biomarker discovery. However, a significant inconsistency in drug sensitivity values among these datasets has been noted. Such inconsistency indicates the presence of substantial noise, subsequently hindering downstream analyses. To address the noise in drug sensitivity data, we introduce a robust and scalable deep learning framework, Residual Thresholded Deep Matrix Factorization (RT-DMF). This method takes a single drug sensitivity data matrix as its sole input and outputs a corrected and imputed matrix. Deep Matrix Factorization (DMF) excels at uncovering subtle patterns, due to its minimal reliance on data structure assumptions. This attribute significantly boosts DMF's ability to identify complex hidden patterns among nuisance effects in the data, thereby facilitating the detection of signals that are therapeutically relevant. Furthermore, RT-DMF incorporates an iterative residual thresholding (RT) procedure, which plays a crucial role in retaining signals more likely to hold therapeutic importance. Validation using simulated datasets and real pharmacogenomics datasets demonstrates the effectiveness of our approach in correcting noise and imputing missing data in drug sensitivity datasets (open source package available at https://github.com/tomwhoooo/rtdmf).

RL2 Enhances the Elimination of Breast Cancer Cells by Doxorubicin

RL2 (recombinant lactaptin 2), a recombinant analogon of the human milk protein Κ-Casein, induces mitophagy and cell death in breast carcinoma cells. Furthermore, RL2 was shown to enhance extrinsic apoptosis upon long-term treatment while inhibiting it upon short-term stimulation. However, the effects of RL2 on the action of chemotherapeutic drugs that induce the intrinsic apoptotic pathway have not been investigated to date. Here, we examined the effects of RL2 on the doxorubicin (DXR)-induced cell death in breast cancer cells with three different backgrounds. In particular, we used BT549 and MDA-MB-231 triple-negative breast cancer (TNBC) cells, T47D estrogen receptor alpha (ERα) positive cells, and SKBR3 human epidermal growth factor receptor 2 (HER2) positive cells. BT549, MDA-MB-231, and T47D cells showed a severe loss of cell viability upon RL2 treatment, accompanied by the induction of mitophagy. Furthermore, BT549, MDA-MB-231, and T47D cells could be sensitized towards DXR treatment with RL2, as evidenced by loss of cell viability. In contrast, SKBR3 cells showed almost no RL2-induced loss of cell viability when treated with RL2 alone, and RL2 did not sensitize SKBR3 cells towards DXR-mediated loss of cell viability. Bioinformatic analysis of gene expression showed an enrichment of genes controlling metabolism in SKBR3 cells compared to the other cell lines. This suggests that the metabolic status of the cells is important for their sensitivity to RL2. Taken together, we have shown that RL2 can enhance the intrinsic apoptotic pathway in TNBC and ERα-positive breast cancer cells, paving the way for the development of novel therapeutic strategies.

The experiences of reproductive concerns in cancer survivors: A systematic review and meta-synthesis of qualitative studies

AIM

The aim of this study was to synthesize qualitative research evidence on cancer survivors' experiences with reproductive concerns (RC).

METHODS

We conducted a systematic search of qualitative studies and utilized the meta-aggregation approach. The database searches were extended up to May 14, 2023, encompassing 12 databases, specifically MEDLINE, CINAHL, PubMed, EMBASE, Scopus, Web of Science (Core Collection), AMED, PsycINFO, The Cochrane Library, CNKI, Wan Fang Data, and VIP.

RESULTS

Three overarching themes were synthesized from the analysis of 21 studies that explored cancer patients' awareness of reproductive concerns, their perceptions, needs, and coping styles. These themes encapsulate the multifaceted aspects of cancer patients' reproductive concerns: "Gender differences in fertility concerns among cancer patients: Perspectives from men and women"; "The influence of age: Experiences with fertility issues among cancer patients at different life stages"; "The impact of treatment stages on fertility concerns: The evolution of perception and coping strategies in the course of cancer treatment".

CONCLUSION

Our study presents an in-depth exploration of the reproductive concerns experienced by cancer patients from various perspectives. We found that the internal experiences of reproductive concerns, their perceptions, needs, and coping mechanisms differ based on their roles. This comprehensive understanding of the complex emotions and needs of cancer patients when confronted with fertility issues can guide clinicians in providing more effective medical assistance, psychological counseling, and fertility-related information services.

Additivity predicts the efficacy of most approved combination therapies for advanced cancer

Most advanced cancers are treated with drug combinations. Rational design aims to identify synergistic combinations, but existing synergy metrics apply to preclinical, not clinical data. Here we propose a model of drug additivity for progression-free survival (PFS) to assess whether clinical efficacies of approved drug combinations are additive or synergistic. This model includes patient-to-patient variability in best single-drug response plus the weaker drug per patient. Among US Food and Drug Administration approvals of drug combinations for advanced cancers (1995-2020), 95% exhibited additive or less than additive effects on PFS times. Among positive or negative phase 3 trials published between 2014-2018, every combination that improved PFS was expected to succeed by additivity (100% sensitivity) and most failures were expected to fail (78% specificity). This study shows synergy is neither a necessary nor common property of clinically effective drug combinations. The predictable efficacy of approved combinations suggests that additivity can be a design principle for combination therapies.