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

Integrative Characterization of Immune-relevant Genes in Hepatocellular Carcinoma

BACKGROUND AND AIMS

Tumor microenvironment plays an essential role in cancer development and progression. Cancer immunotherapy has become a promising approach for the treatment of hepatocellular carcinoma (HCC). We aimed to analyze the HCC immune microenvironment characteristics to identify immune-related genetic changes.

METHODS

Key immune-relevant genes (KIRGs) were obtained through integrating the differentially expressed genes of The Cancer Genome Atlas, immune genes from the Immunology Database and Analysis Portal, and immune differentially expressed genes determined by single-sample gene set enrichment analysis scores. Cox regression analysis was performed to mine therapeutic target genes. A regulatory network based on KIRGs, transcription factors, and immune-related long non-coding RNAs (IRLncRNAs) was also generated. The outcomes of risk score model were validated in a testing cohort and in clinical samples using tissue immunohistochemistry staining. Correlation analysis between risk score and immune checkpoint genes and immune cell infiltration were investigated.

RESULTS

In total, we identified 21 KIRGs, including programmed cell death-1 (PD-1) and cytotoxic T-lymphocyte associated protein 4 (CTLA4), and found IKBKE, IL2RG, EDNRA, and IGHA1 may be equally important to PD-1 or CTLA4. Meanwhile, KIRGs, various transcription factors, and IRLncRNAs were integrated to reveal that the NRF1-AC127024.5-IKBKE axis might be involved in tumor immunity regulation. Furthermore, the immune-related risk score model was established according to KIRGs and key IRLncRNAs, and verified more obvious discriminating power in the testing cohort. Correlation analysis indicated TNFSF4 , LGALS9 , KIAA1429 , IDO2, and CD276 were closely related to the risk score, and CD4 T cells, macrophages, and neutrophils were the primary immune infiltration cell types.

CONCLUSIONS

Our results highlight the importance of immune genes in the HCC microenvironment and further unravel the underlying molecular mechanisms in the development of HCC.

Opportunities for Ferroptosis in Cancer Therapy

A critical hallmark of cancer cells is their ability to evade programmed apoptotic cell death. Consequently, resistance to anti-cancer therapeutics is a hurdle often observed in the clinic. Ferroptosis, a non-apoptotic form of cell death distinguished by toxic lipid peroxidation and iron accumulation, has garnered substantial attention as an alternative therapeutic strategy to selectively destroy tumours. Although there is a plethora of research outlining the molecular mechanisms of ferroptosis, these findings are yet to be translated into clinical compounds inducing ferroptosis. In this perspective, we elaborate on how ferroptosis can be leveraged in the clinic. We discuss a therapeutic window for compounds inducing ferroptosis, the subset of tumour types that are most sensitive to ferroptosis, conventional therapeutics that induce ferroptosis, and potential strategies for lowering the threshold for ferroptosis.

The Use of Informer Sets in Screening: Perspectives on an Efficient Strategy to Identify New Probes

Small-molecule discovery typically involves large-scale screening campaigns, spanning multiple compound collections. However, such activities can be cost- or time-prohibitive, especially when using complex assay systems, limiting the number of compounds tested. Further, low hit rates can make the process inefficient. Sparse coverage of chemical structure or biological activity space can lead to limited success in a primary screen and represents a missed opportunity by virtue of selecting the "wrong" compounds to test. Thus, the choice of screening collections becomes of paramount importance. In this perspective, we discuss the utility of generating "informer sets" for small-molecule discovery, and how this strategy can be leveraged to prioritize probe candidates. While many researchers may assume that informer sets are focused on particular targets (e.g., kinases) or processes (e.g., autophagy), efforts to assemble informer sets based on historical bioactivity or successful human exposure (e.g., repurposing collections) have shown promise as well. Another method for generating informer sets is based on chemical structure, particularly when the compounds have unknown activities and targets. We describe our efforts to screen an informer set representing a collection of 100,000 small molecules synthesized through diversity-oriented synthesis (DOS). This process enables researchers to identify activity early and more extensively screen only a few chemical scaffolds, rather than the entire collection. This elegant and economic outcome is a goal of the informer set approach. Here, we aim not only to shed light on this process, but also to promote the use of informer sets more widely in small-molecule discovery projects.

Minimal information for Chemosensitivity assays (MICHA): A next-generation pipeline to enable the FAIRification of drug screening experiments

Chemosensitivity assays are commonly used for preclinical drug discovery and clinical trial optimization. However, data from independent assays are often discordant, largely attributed to uncharacterized variation in the experimental materials and protocols. We report here the launching of MICHA (Minimal Information for Chemosensitivity Assays), accessed via https://micha-protocol.org. Distinguished from existing efforts that are often lacking support from data integration tools, MICHA can automatically extract publicly available information to facilitate the assay annotation including: 1) compounds, 2) samples, 3) reagents, and 4) data processing methods. For example, MICHA provides an integrative web server and database to obtain compound annotation including chemical structures, targets, and disease indications. In addition, the annotation of cell line samples, assay protocols and literature references can be greatly eased by retrieving manually curated catalogues. Once the annotation is complete, MICHA can export a report that conforms to the FAIR principle (Findable, Accessible, Interoperable and Reusable) of drug screening studies. To consolidate the utility of MICHA, we provide FAIRified protocols from five major cancer drug screening studies, as well as six recently conducted COVID-19 studies. With the MICHA webserver and database, we envisage a wider adoption of a community-driven effort to improve the open access of drug sensitivity assays.

Integrative Transcriptomic Analysis Reveals Distinctive Molecular Traits and Novel Subtypes of Collecting Duct Carcinoma

Collecting duct carcinoma (CDC) is a rare and highly aggressive kidney cancer subtype with poor prognosis and no standard treatments. To date, only a few studies have examined the transcriptomic portrait of CDC. Through integration of multiple datasets, we compared CDC to normal tissue, upper-tract urothelial carcinomas, and other renal cancers, including clear cell, papillary, and chromophobe histologies. Association between CDC gene expression signatures and in vitro drug sensitivity data was evaluated using the Cancer Therapeutic Response Portal, Genomics of Drug Sensitivity in Cancer datasets, and connectivity map. We identified a CDC-specific gene signature that predicted in vitro sensitivity to different targeted agents and was associated to worse outcome in clear cell renal cell carcinoma. We showed that CDC are transcriptionally related to the principal cells of the collecting ducts providing evidence that this tumor originates from this normal kidney cell type. Finally, we proved that CDC is a molecularly heterogeneous disease composed of at least two subtypes distinguished by cell signaling, metabolic and immune-related alterations. Our findings elucidate the molecular features of CDC providing novel biological and clinical insights. The identification of distinct CDC subtypes and their transcriptomic traits provides the rationale for patient stratification and alternative therapeutic approaches.

Anti-tumor effects of rigosertib in high-risk neuroblastoma

High-risk neuroblastoma has a poor prognosis despite intense treatment, demonstrating the need for new therapeutic strategies. Here we evaluated the effects of rigosertib (ON-01910.Na) in preclinical models of high-risk neuroblastoma. Among several hundred cancer cell lines representing 24 tumor types, neuroblastoma was the most sensitive to rigosertib. Treatment of MYCN-amplified neuroblastoma organoids resulted in organoid disintegration, decreased cell viability, and increased apoptotic cell death. Neuroblastoma response to rigosertib involved G2M cell cycle arrest and decreased phosphorylation of AKT (Ser473) and ERK1/2 (Thr202/Tyr204). Rigosertib delayed tumor growth and prolonged survival of mice carrying neuroblastoma MYCN-amplified PDX tumors (median survival: 31 days, treated; 22 days, vehicle) accompanied with increased apoptosis in treated tumors. We further identified vincristine and rigosertib as a potential promising drug combination treatment. Our results show that rigosertib might be a useful therapeutic agent for MYCN-amplified neuroblastomas, especially in combination with existing agents.

A Pan-cancer Analysis Reveals the Abnormal Expression and Drug Sensitivity of CSF1

BACKGROUND

Colony-stimulating factor-1 (CSF1) is a cytokine that is closely related to normal organ growth and development as well as tumor progression.

OBJECTIVE

We aimed to summarize and clarify the reasons for the abnormal expression of CSF1 in tumors and explore the role of CSF1 in tumor progression. Furthermore, drug response analysis may provide a reference for clinical medication.

METHODS

The expression of CSF1 was analyzed by TCGA and CCLE. Besides, cBioPortal and MethSurv databases were used to conduct mutation and DNA methylation analyses. Further, correlations between CSF1 expression and tumor stage, survival, immune infiltration, drug sensitivity and enrichment analyses were validated via UALCAN, Kaplan-Meier plotter, TIMER, CTRP and Coexperia databases.

RESULTS

CSF1 is expressed in a variety of tissues, meaningfully, it can be detected in blood. Compared with normal tissues, CSF1 expression was significantly decreased in most tumors. The missense mutation and DNA methylation of CSF1 may cause the downregulated expression. Moreover, decreased CSF1 expression was related with higher tumor stage and worse survival. Further, the promoter DNA methylation level of CSF1 was prognostically significant in most tumors. Besides, CSF1 was closely related to immune infiltration, especially macrophages. Importantly, CSF1 expression was associated with a good response to VEGFRs inhibitors, which may be due to the possible involvement of CSF1 in tumor angiogenesis and metastasis processes.

CONCLUSION

The abnormal expression of CSF1 could serve as a promising biomarker of tumor progression and prognosis in pan-cancer. Significantly, angiogenesis and metastasis inhibitors may show a good response to CSF1-related tumors.

Preclinical models as patients' avatars for precision medicine in colorectal cancer: past and future challenges

Colorectal cancer (CRC) is a complex and heterogeneous disease, characterized by dismal prognosis and low survival rate in the advanced (metastatic) stage. During the last decade, the establishment of novel preclinical models, leading to the generation of translational discovery and validation platforms, has opened up a new scenario for the clinical practice of CRC patients. To bridge the results developed at the bench with the medical decision process, the ideal model should be easily scalable, reliable to predict treatment responses, and flexibly adapted for various applications in the research. As such, the improved benefit of novel therapies being tested initially on valuable and reproducible preclinical models would lie in personalized treatment recommendations based on the biology and genomics of the patient's tumor with the overall aim to avoid overtreatment and unnecessary toxicity. In this review, we summarize different in vitro and in vivo models, which proved efficacy in detection of novel CRC culprits and shed light into the biology and therapy of this complex disease. Even though cell lines and patient-derived xenografts remain the mainstay of colorectal cancer research, the field has been confidently shifting to the use of organoids as the most relevant preclinical model. Prioritization of organoids is supported by increasing body of evidence that these represent excellent tools worth further therapeutic explorations. In addition, novel preclinical models such as zebrafish avatars are emerging as useful tools for pharmacological interrogation. Finally, all available models represent complementary tools that can be utilized for precision medicine applications.

Assessing Consistency Across Functional Screening Datasets in Cancer Cells

Motivation

Many high-throughput screening studies have been carried out in cancer cell lines to identify therapeutic agents and targets. Existing consistency assessment studies only examined two datasets at a time, with conclusions based on a subset of carefully selected features rather than considering global consistency of all the data. However, poor concordance can still be observed for a large part of the data even when selected features are highly consistent.

Results

In this study, we assembled nine compound screening datasets and three functional genomics datasets. We derived direct measures of consistency as well as indirect measures of consistency based on association between functional data and copy number-adjusted gene expression data. These results have been integrated into a web application—the Functional Data Consistency Explorer (FDCE), to allow users to make queries and generate interactive visualizations so that functional data consistency can be assessed for individual features of interest.

Availability and implementation

The FDCE web tool and we have developed and the functional data consistency measures we have generated are available at https://lccl.shinyapps.io/FDCE/.

Supplementary information

Supplementary data are available at Bioinformatics online.

PRoBE the cloud toolkit: finding the best biomarkers of drug response within a breast cancer clinical trial

Objectives

In this paper, we discuss leveraging cloud-based platforms to collect, visualize, analyze, and share data in the context of a clinical trial. Our cloud-based infrastructure, Patient Repository of Biomolecular Entities (PRoBE), has given us the opportunity for uniform data structure, more efficient analysis of valuable data, and increased collaboration between researchers.

Materials and Methods

We utilize a multi-cloud platform to manage and analyze data generated from the clinical Investigation of Serial Studies to Predict Your Therapeutic Response with Imaging And moLecular Analysis 2 (I-SPY 2 TRIAL). A collaboration with the Institute for Systems Biology Cancer Gateway in the Cloud has additionally given us access to public genomic databases. Applications to I-SPY 2 data have been built using R Shiny, while leveraging Google's BigQuery tables and SQL commands for data mining.

Results

We highlight the implementation of PRoBE in several unique case studies including prediction of biomarkers associated with clinical response, access to the Pan-Cancer Atlas, and integrating pathology images within the cloud. Our data integration pipelines, documentation, and all codebase will be placed in a Github repository.

Discussion and conclusion

We are hoping to develop risk stratification diagnostics by integrating additional molecular, magnetic resonance imaging, and pathology markers into PRoBE to better predict drug response. A robust cloud infrastructure and tool set can help integrate these large datasets to make valuable predictions of response to multiple agents. For that reason, we are continuously improving PRoBE to advance the way data is stored, accessed, and analyzed in the I-SPY 2 clinical trial.

Differential Outcomes in Codon 12/13 and Codon 61 NRAS-Mutated Cancers in the Phase II NCI-MATCH Trial of Binimetinib in Patients with NRAS-Mutated Tumors

PURPOSE

Preclinical and clinical data suggest that downstream inhibition with an MEK inhibitor, such as binimetinib, might be efficacious for NRAS-mutated cancers.

PATIENTS AND METHODS

Patients enrolled in the NCI-MATCH trial master protocol underwent tumor biopsy and molecular profiling by targeted next-generation sequencing. Patients with NRAS-mutated tumors, except melanoma, were enrolled in subprotocol Z1A, a single-arm study evaluating binimetinib 45 mg twice daily. The primary endpoint was objective response rate (ORR). Secondary endpoints included progression-free survival (PFS) and overall survival (OS). A post hoc analysis examined the association of NRAS mutation type with outcome.

RESULTS

In total, 47 eligible patients with a refractory solid tumor harboring a codon 12, 13, or 61 NRAS mutation were treated. Observed toxicity was moderate, and 30% of patients discontinued treatment because of binimetinib-associated toxicity. The ORR was 2.1% (1/47 patients). A patient with malignant ameloblastoma harboring a codon 61 NRAS mutation achieved a durable partial response (PR). A patient with NRAS codon 61-mutated colorectal cancer had an unconfirmed PR, and two other patients with NRAS codon 61-mutated colorectal had stable disease for at least 12 months. In an exploratory analysis, patients with colorectal cancer bearing a NRAS codon 61 mutation (n = 8) had a significantly longer OS (P = 0.03) and PFS (P = 0.007) than those with codon 12 or 13 mutations (n = 16).

CONCLUSIONS

Single-agent binimetinib did not show promising efficacy in NRAS-mutated cancers. The observation of increased OS and PFS in patients with codon 61 NRAS-mutated colorectal cancer merits further investigation.

Identification of inhibitory immune checkpoints and relevant regulatory pathways in breast cancer stem cells

Immune checkpoint blockade (ICB) has become one of the most promising approaches to activating antitumor immunity. However, only a small subset of patients with breast cancer benefit from ICB treatment. To improve the therapeutic effect in the clinic, precision immunotherapy is proposed to accurately eliminate cancer stem cells that contribute to local recurrence or metastasis, but currently little is known about their immunological properties. In this study, breast cancer-specific datasets in The Cancer Genome Atlas were collected and analyzed by using multiple open-access web servers. We found that the immunophenotype of breast cancer was characterized by a hypoactive immune microenvironment and a low response to immune checkpoint blockade. The innate immune checkpoint CD200 and the adaptive immune checkpoint CD276, respectively, exhibited a strong correlation with basal/stem gene signature and invasiveness gene signature, both of which represent breast cancer stem cells. Wnt, TGF-β, and Hedgehog signaling, which are recognized as stemness-related pathways, showed a significant association with the expression of CD200 and CD276, suggesting cancer stem cell-specific immune checkpoints could be downregulated by inhibiting these pathways. Of note, levels of CD200 and CD276 expression were higher in TGF-β dominant breast cancer than in other immune types of breast cancer. We also identified gene signatures that represent Wnt, TGF-β, and Hedgehog signaling-related CD200 and CD276 expression in breast cancer stem cells. For the luminal A subtype, the patient group with a high level of these gene signatures plus a low infiltration of CD8+ T cells, or dendritic cells, or M1 macrophages had poor overall survival. Our study suggested that CD200 and CD276 are candidate inhibitory immune checkpoints in breast cancer stem cells, which are potentially regulated by Wnt, TGF-β, and Hedgehog signaling. Synergistic inhibition of these stemness-related pathways may improve the efficacy of ICB treatment targeting breast cancer stem cells in precision immunotherapy.

kESVR: An Ensemble Model for Drug Response Prediction in Precision Medicine Using Cancer Cell Lines Gene Expression

Background: Cancer cell lines are frequently used in research as in-vitro tumor models. Genomic data and large-scale drug screening have accelerated the right drug selection for cancer patients. Accuracy in drug response prediction is crucial for success. Due to data-type diversity and big data volume, few methods can integrative and efficiently find the principal low-dimensional manifold of the high-dimensional cancer multi-omics data to predict drug response in precision medicine. Method: A novelty k-means Ensemble Support Vector Regression (kESVR) is developed to predict each drug response values for single patient based on cell-line gene expression data. The kESVR is a blend of supervised and unsupervised learning methods and is entirely data driven. It utilizes embedded clustering (Principal Component Analysis and k-means clustering) and local regression (Support Vector Regression) to predict drug response and obtain the global pattern while overcoming missing data and outliers’ noise. Results: We compared the efficiency and accuracy of kESVR to 4 standard machine learning regression models: (1) simple linear regression, (2) support vector regression (3) random forest (quantile regression forest) and (4) back propagation neural network. Our results, which based on drug response across 610 cancer cells from Cancer Cell Line Encyclopedia (CCLE) and Cancer Therapeutics Response Portal (CTRP v2), proved to have the highest accuracy (smallest mean squared error (MSE) measure). We next compared kESVR with existing 17 drug response prediction models based a varied range of methods such as regression, Bayesian inference, matrix factorization and deep learning. After ranking the 18 models based on their accuracy of prediction, kESVR ranks first (best performing) in majority (74%) of the time. As for the remaining (26%) cases, kESVR still ranked in the top five performing models. Conclusion: In this paper we introduce a novel model (kESVR) for drug response prediction using high dimensional cell-line gene expression data. This model outperforms current existing prediction models in terms of prediction accuracy and speed and overcomes overfitting. This can be used in future to develop a robust drug response prediction system for cancer patients using the cancer cell-lines guidance and multi-omics data.

Identification of biological targets through the correlation between cell line chemosensitivity and protein expression pattern

Matching biological data sequences is one of the most interesting ways to discover new bioactive compounds. In particular, matching cell chemosensitivity with a protein expression profile can be a useful approach to predict the activity of compounds against definite biological targets. In this review, we discuss this correlation. First, we analyze case studies in which some known drugs, acting on known targets, show a good correlation between their antiproliferative activities and protein expression when a large panel of tumor cells is considered. Then, we highlight how the application of in silico methods based on the correlation between cell line chemosensitivity and gene/protein expression patterns might be a quick, cheap, and interesting approach to predict the biological activity of investigated molecules.

FOXM1 repression increases mitotic death upon antimitotic chemotherapy through BMF upregulation

Inhibition of spindle microtubule (MT) dynamics has been effectively used in cancer treatment. Although the mechanisms by which MT poisons elicit mitotic arrest are fairly understood, efforts are still needed towards elucidating how cancer cells respond to antimitotic drugs owing to cytotoxicity and resistance side effects. Here, we identified the critical G2/M transcription factor Forkhead box M1 (FOXM1) as a molecular determinant of cell response to antimitotics. We found FOXM1 repression to increase death in mitosis (DiM) due to upregulation of the BCL-2 modifying factor (BMF) gene involved in anoikis, an apoptotic process induced upon cell detachment from the extracellular matrix. FOXM1 binds to a BMF intronic cis-regulatory element that interacts with both the BMF and the neighbor gene BUB1B promoter regions, to oppositely regulate their expression. This mechanism ensures that cells treated with antimitotics repress BMF and avoid DiM when FOXM1 levels are high. In addition, we show that this mechanism is partly disrupted in anoikis/antimitotics-resistant tumor cells, with resistance correlating with lower BMF expression but in a FOXM1-independent manner. These findings provide a stratification biomarker for antimitotic chemotherapy response.

Super.FELT: supervised feature extraction learning using triplet loss for drug response prediction with multi-omics data

BACKGROUND

Predicting the drug response of a patient is important for precision oncology. In recent studies, multi-omics data have been used to improve the prediction accuracy of drug response. Although multi-omics data are good resources for drug response prediction, the large dimension of data tends to hinder performance improvement. In this study, we aimed to develop a new method, which can effectively reduce the large dimension of data, based on the supervised deep learning model for predicting drug response.

RESULTS

We proposed a novel method called Supervised Feature Extraction Learning using Triplet loss (Super.FELT) for drug response prediction. Super.FELT consists of three stages, namely, feature selection, feature encoding using a supervised method, and binary classification of drug response (sensitive or resistant). We used multi-omics data including mutation, copy number aberration, and gene expression, and these were obtained from cell lines [Genomics of Drug Sensitivity in Cancer (GDSC), Cancer Cell Line Encyclopedia (CCLE), and Cancer Therapeutics Response Portal (CTRP)], patient-derived tumor xenografts (PDX), and The Cancer Genome Atlas (TCGA). GDSC was used for training and cross-validation tests, and CCLE, CTRP, PDX, and TCGA were used for external validation. We performed ablation studies for the three stages and verified that the use of multi-omics data guarantees better performance of drug response prediction. Our results verified that Super.FELT outperformed the other methods at external validation on PDX and TCGA and was good at cross-validation on GDSC and external validation on CCLE and CTRP. In addition, through our experiments, we confirmed that using multi-omics data is useful for external non-cell line data.

CONCLUSION

By separating the three stages, Super.FELT achieved better performance than the other methods. Through our results, we found that it is important to train encoders and a classifier independently, especially for external test on PDX and TCGA. Moreover, although gene expression is the most powerful data on cell line data, multi-omics promises better performance for external validation on non-cell line data than gene expression data. Source codes of Super.FELT are available at  https://github.com/DMCB-GIST/Super.FELT .

Drug Repurposing in Oncology, an Attractive Opportunity for Novel Combinatorial Regimens

The costs of developing, validating and buying new drugs are dramatically increasing. On the other hand, sobering economies have difficulties in sustaining their healthcare systems, particularly in countries with an elderly population requiring increasing welfare. This conundrum requires immediate action, and a possible option is to study the large, already present arsenal of drugs approved and to use them for innovative therapies. This possibility is particularly interesting in oncology, where the complexity of the cancer genome dictates in most patients a multistep therapeutic approach. In this review, we discuss a) Computational approaches; b) preclinical models; c) currently ongoing or already published clinical trials in the drug repurposing field in oncology; and d) drug repurposing to overcome resistance to previous therapies.

Current Advances and Limitations of Deep Learning in Anticancer Drug Sensitivity Prediction

Anticancer drug screening can accelerate drug discovery to save the lives of cancer patients, but cancer heterogeneity makes this screening challenging. The prediction of anticancer drug sensitivity is useful for anticancer drug development and the identification of biomarkers of drug sensitivity. Deep learning, as a branch of machine learning, is an important aspect of in silico research. Its outstanding computational performance means that it has been used for many biomedical purposes, such as medical image interpretation, biological sequence analysis, and drug discovery. Several studies have predicted anticancer drug sensitivity based on deep learning algorithms. The field of deep learning has made progress regarding model performance and multi-omics data integration. However, deep learning is limited by the number of studies performed and data sources available, so it is not perfect as a pre-clinical approach for use in the anticancer drug screening process. Improving the performance of deep learning models is a pressing issue for researchers. In this review, we introduce the research of anticancer drug sensitivity prediction and the use of deep learning in this research area. To provide a reference for future research, we also review some common data sources and machine learning methods. Lastly, we discuss the advantages and disadvantages of deep learning, as well as the limitations and future perspectives regarding this approach.

Learning curves for drug response prediction in cancer cell lines

BACKGROUND

Motivated by the size and availability of cell line drug sensitivity data, researchers have been developing machine learning (ML) models for predicting drug response to advance cancer treatment. As drug sensitivity studies continue generating drug response data, a common question is whether the generalization performance of existing prediction models can be further improved with more training data.

METHODS

We utilize empirical learning curves for evaluating and comparing the data scaling properties of two neural networks (NNs) and two gradient boosting decision tree (GBDT) models trained on four cell line drug screening datasets. The learning curves are accurately fitted to a power law model, providing a framework for assessing the data scaling behavior of these models.

RESULTS

The curves demonstrate that no single model dominates in terms of prediction performance across all datasets and training sizes, thus suggesting that the actual shape of these curves depends on the unique pair of an ML model and a dataset. The multi-input NN (mNN), in which gene expressions of cancer cells and molecular drug descriptors are input into separate subnetworks, outperforms a single-input NN (sNN), where the cell and drug features are concatenated for the input layer. In contrast, a GBDT with hyperparameter tuning exhibits superior performance as compared with both NNs at the lower range of training set sizes for two of the tested datasets, whereas the mNN consistently performs better at the higher range of training sizes. Moreover, the trajectory of the curves suggests that increasing the sample size is expected to further improve prediction scores of both NNs. These observations demonstrate the benefit of using learning curves to evaluate prediction models, providing a broader perspective on the overall data scaling characteristics.

CONCLUSIONS

A fitted power law learning curve provides a forward-looking metric for analyzing prediction performance and can serve as a co-design tool to guide experimental biologists and computational scientists in the design of future experiments in prospective research studies.

Pharmacological or TRIB3-Mediated Suppression of ATF4 Transcriptional Activity Promotes Hepatoma Cell Resistance to Proteasome Inhibitor Bortezomib

Simple Summary

Proteasome inhibitors are currently used in the treatment of certain blood cancers, and clinical trials to treat solid tumors, including liver cancer, have also been conducted. However, different malignancies are not equally susceptible to proteasome inhibitors, and resistance to the drug may develop during the therapy. Here, we characterize the molecular mechanisms underlying the resilience of liver cancer cells to the proteasome inhibitor bortezomib. The results demonstrate that the activity of the eIF2α–ATF4 stress response pathway affects the viability of cells treated with bortezomib. We found that the pseudokinase TRIB3, an endogenous regulator of ATF4 and a gene highly expressed in liver cancer, resides predominantly at the same chromatin sites as ATF4 and constrains ATF4 activity. The survival of bortezomib-exposed hepatoma cells proved sensitive to TRIB3 overexpression and inactivation. Thus, TRIB3 is a novel factor contributing to bortezomib resistance of liver cancer cells.

Abstract

The proteasome is an appealing target for anticancer therapy and the proteasome inhibitor bortezomib has been approved for the treatment of several types of malignancies. However, the molecular mechanisms underlying cancer cell resistance to bortezomib remain poorly understood. In the current article, we investigate how modulation of the eIF2α–ATF4 stress pathway affects hepatoma cell response to bortezomib. Transcriptome profiling revealed that many ATF4 transcriptional target genes are among the most upregulated genes in bortezomib-treated HepG2 human hepatoma cells. While pharmacological enhancement of the eIF2α–ATF4 pathway activity results in the elevation of the activities of all branches of the unfolded protein response (UPR) and sensitizes cells to bortezomib toxicity, the suppression of ATF4 induction delays bortezomib-induced cell death. The pseudokinase TRIB3, an inhibitor of ATF4, is expressed at a high basal level in hepatoma cells and is strongly upregulated in response to bortezomib. To map genome-wide chromatin binding loci of TRIB3 protein, we fused a Flag tag to endogenous TRIB3 in HepG2 cells and performed ChIP-Seq. The results demonstrate that TRIB3 predominantly colocalizes with ATF4 on chromatin and binds to genomic regions containing the C/EBP–ATF motif. Bortezomib treatment leads to a robust enrichment of TRIB3 binding near genes induced by bortezomib and involved in the ER stress response and cell death. Disruption of TRIB3 increases C/EBP–ATF-driven transcription, augments ER stress and cell death upon exposure to bortezomib, while TRIB3 overexpression enhances cell survival. Thus, TRIB3, colocalizing with ATF4 and limiting its transcriptional activity, functions as a factor increasing resistance to bortezomib, while pharmacological over-activation of eIF2α–ATF4 can overcome the endogenous restraint mechanisms and sensitize cells to bortezomib.

High BRCA2 Gene Expression is Associated with Aggressive and Highly Proliferative Breast Cancer

BACKGROUND

Mutations of BRCA genes are the most studied in breast cancer, but the clinical relevance of BRCA2 gene expression has been less well studied. Given that BRCA2 is a DNA repair gene, we hypothesized that high BRCA2 expression is associated with highly proliferative and aggressive biology in breast cancer.

MATERIALS AND METHODS

A total of 4342 breast cancer patients were analyzed from The Cancer Genome Atlas (TCGA, n = 1069) as the testing cohort and Gene Expression Omnibus (GEO) dataset GSE96058 (n = 3273) as a validation cohort.

RESULTS

There was no relationship between BRCA2 mutation and BRCA2 gene expression. BRCA2 high expression breast cancer was associated with higher Nottingham grade (p < 0.001), with high proliferation (MKI-67, p < 0.001), and with higher intratumor heterogeneity, homologous recombination deficiency, mutation rate, fraction altered, and neoantigens (all p < 0.001). BRCA2 high expression was associated with E2F1, RB1, PALB2, and PARP, as well as cell proliferation-related gene sets, E2F targets, G2M checkpoints, and mitotic spindle, and with less ESR1 and ER response early and late gene sets. They were associated with significantly reduced number of stroma cells and with high infiltration of both favorable and unfavorable immune cells. BRCA2 high expression significantly correlated with response to olaparib, a PARP inhibitor, and inversely with cyclophosphamide in ER-positive/HER2-negative tumors, but not in TNBC.

CONCLUSIONS

BRCA2 high gene expression was associated with highly proliferative and aggressive breast cancer that was highly immunogenic with better response to PARP inhibitors in ER-positive patients. BRCA2 gene expression may become a candidate marker for aggressive biology and to tailor specific treatment strategies in the future.

Detection of Molecular Signatures of Homologous Recombination Deficiency in Bladder Cancer

PURPOSE

Poly (ADP ribose)-polymerase (PARP) inhibitors are approved for use in breast, ovarian, prostate, and pancreatic cancers, which are the solid tumor types that most frequently have alterations in key homologous recombination (HR) genes, such as BRCA1/2. However, the frequency of HR deficiency (HRD) in other solid tumor types, including bladder cancer, is less well characterized.

EXPERIMENTAL DESIGN

Specific DNA aberration profiles (mutational signatures) are induced by HRD, and the presence of these "genomic scars" can be used to assess the presence or absence of HRD in a given tumor biopsy even in the absence of an observed alteration of an HR gene. Using whole-exome and whole-genome data, we measured various HRD-associated mutational signatures in bladder cancer.

RESULTS

We found that a subset of bladder tumors have evidence of HRD. In addition to a small number of tumors with biallelic BRCA1/2 events, approximately 10% of bladder tumors had significant evidence of HRD-associated mutational signatures. Increased levels of HRD signatures were associated with promoter methylation of RBBP8, which encodes CtIP, a key protein involved in HR.

CONCLUSIONS

A subset of bladder tumors have genomic features suggestive of HRD and therefore may be more likely to benefit from therapies such as platinum agents and PARP inhibitors that target tumor HRD.

Systematic Characterization of Recurrent Genomic Alterations in Cyclin-Dependent Kinases Reveals Potential Therapeutic Strategies for Cancer Treatment

Recurrent copy-number alterations, mutations, and transcript fusions of the genes encoding CDKs/cyclins are characterized in >10,000 tumors. Genomic alterations of CDKs/cyclins are dominantly driven by copy number aberrations. In contrast to cell-cycle-related CDKs/cyclins, which are globally amplified, transcriptional CDKs/cyclins recurrently lose copy numbers across cancers. Although mutations and transcript fusions are relatively rare events, CDK12 exhibits recurrent mutations in multiple cancers. Among the transcriptional CDKs, CDK7 and CDK12 show the most significant copy number loss and mutation, respectively. Their genomic alterations are correlated with increased sensitivities to DNA-damaging drugs. Inhibition of CDK7 preferentially represses the expression of genes in the DNA-damage-repair pathways and impairs the activity of homologous recombination. Low-dose CDK7 inhibitor treatment sensitizes cancer cells to PARP inhibitor-induced DNA damage and cell death. Our analysis provides genomic information for identification and prioritization of drug targets for CDKs and reveals rationales for treatment strategies.

A Bioinformatics Analysis Identifies the Telomerase Inhibitor MST-312 for Treating High-STMN1-Expressing Hepatocellular Carcinoma

Hepatocellular carcinoma (HCC) is a relatively chemo-resistant tumor. Several multi-kinase inhibitors have been approved for treating advanced HCC. However, most HCC patients are highly refractory to these drugs. Therefore, the development of more effective therapies for advanced HCC patients is urgently needed. Stathmin 1 (STMN1) is an oncoprotein that destabilizes microtubules and promotes cancer cell migration and invasion. In this study, cancer genomics data mining identified STMN1 as a prognosis biomarker and a therapeutic target for HCC. Co-expressed gene analysis indicated that STMN1 expression was positively associated with cell-cycle-related gene expression. Chemical sensitivity profiling of HCC cell lines suggested that High-STMN1-expressing HCC cells were the most sensitive to MST-312 (a telomerase inhibitor). Drug-gene connectivity mapping supported that MST-312 reversed the STMN1-co-expressed gene signature (especially BUB1B, MCM2/5/6, and TTK genes). In vitro experiments validated that MST-312 inhibited HCC cell viability and related protein expression (STMN1, BUB1B, and MCM5). In addition, overexpression of STMN1 enhanced the anticancer activity of MST-312 in HCC cells. Therefore, MST-312 can be used for treating STMN1-high expression HCC.

Predicting tumor response to drugs based on gene-expression biomarkers of sensitivity learned from cancer cell lines

Background

Human cancer cell line profiling and drug sensitivity studies provide valuable information about the therapeutic potential of drugs and their possible mechanisms of action. The goal of those studies is to translate the findings from in vitro studies of cancer cell lines into in vivo therapeutic relevance and, eventually, patients’ care. Tremendous progress has been made.

Results

In this work, we built predictive models for 453 drugs using data on gene expression and drug sensitivity (IC₅₀) from cancer cell lines. We identified many known drug-gene interactions and uncovered several potentially novel drug-gene associations. Importantly, we further applied these predictive models to ~ 17,000 bulk RNA-seq samples from The Cancer Genome Atlas (TCGA) and the Genotype-Tissue Expression (GTEx) database to predict drug sensitivity for both normal and tumor tissues. We created a web site for users to visualize and download our predicted data (https://manticore.niehs.nih.gov/cancerRxTissue). Using trametinib as an example, we showed that our approach can faithfully recapitulate the known tumor specificity of the drug.

Conclusions

We demonstrated that our approach can predict drugs that 1) are tumor-type specific; 2) elicit higher sensitivity from tumor compared to corresponding normal tissue; 3) elicit differential sensitivity across breast cancer subtypes. If validated, our prediction could have relevance for preclinical drug testing and in phase I clinical design.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12864-021-07581-7.

Targeting pan-essential genes in cancer: Challenges and opportunities

Despite remarkable successes in the clinic, cancer targeted therapy development remains challenging and the failure rate is disappointingly high. This problem is partly due to the misapplication of the targeted therapy paradigm to therapeutics targeting pan-essential genes, which can result in therapeutics whereby efficacy is attenuated by dose-limiting toxicity. Here we summarize the key features of successful chemotherapy and targeted therapy agents, and use case studies to outline recurrent challenges to drug development efforts targeting pan-essential genes. Finally, we suggest strategies to avoid previous pitfalls for ongoing and future development of pan-essential therapeutics.

Leveraging Single-Cell Approaches in Cancer Precision Medicine

Cancer precision medicine aims to improve patient outcomes by tailoring treatment to the unique genomic background of a tumor. However, efforts to develop prognostic and drug response biomarkers largely rely on bulk 'omic' data, which fails to capture intratumor heterogeneity (ITH) and deconvolve signals from normal versus tumor cells. These shortcomings in measuring clinically relevant features are being addressed with single-cell technologies, which provide a fine-resolution map of the genetic and phenotypic heterogeneity in tumors and their microenvironment, as well as an improved understanding of the patterns of subclonal tumor populations. Here we present recent advances in the application of single-cell technologies, towards gaining a deeper understanding of ITH and evolution, and potential applications in developing personalized therapeutic strategies.

The right tools for the job: the central role for next generation chemical probes and chemistry-based target deconvolution methods in phenotypic drug discovery

The reconnection of the scientific community with phenotypic drug discovery has created exciting new possibilities to develop therapies for diseases with highly complex biology. It promises to revolutionise fields such as neurodegenerative disease and regenerative medicine, where the development of new drugs has consistently proved elusive. Arguably, the greatest challenge in readopting the phenotypic drug discovery approach exists in establishing a crucial chain of translatability between phenotype and benefit to patients in the clinic. This remains a key stumbling block for the field which needs to be overcome in order to fully realise the potential of phenotypic drug discovery. Excellent quality chemical probes and chemistry-based target deconvolution techniques will be a crucial part of this process. In this review, we discuss the current capabilities of chemical probes and chemistry-based target deconvolution methods and evaluate the next advances necessary in order to fully support phenotypic screening approaches in drug discovery.

Exploration of databases and methods supporting drug repurposing: a comprehensive survey

Drug development involves a deep understanding of the mechanisms of action and possible side effects of each drug, and sometimes results in the identification of new and unexpected uses for drugs, termed as drug repurposing. Both in case of serendipitous observations and systematic mechanistic explorations, confirmation of new indications for a drug requires hypothesis building around relevant drug-related data, such as molecular targets involved, and patient and cellular responses. These datasets are available in public repositories, but apart from sifting through the sheer amount of data imposing computational bottleneck, a major challenge is the difficulty in selecting which databases to use from an increasingly large number of available databases. The database selection is made harder by the lack of an overview of the types of data offered in each database. In order to alleviate these problems and to guide the end user through the drug repurposing efforts, we provide here a survey of 102 of the most promising and drug-relevant databases reported to date. We summarize the target coverage and types of data available in each database and provide several examples of how multi-database exploration can facilitate drug repurposing.

mTORC1 couples cyst(e)ine availability with GPX4 protein synthesis and ferroptosis regulation

Glutathione peroxidase 4 (GPX4) utilizes glutathione (GSH) to detoxify lipid peroxidation and plays an essential role in inhibiting ferroptosis. As a selenoprotein, GPX4 protein synthesis is highly inefficient and energetically costly. How cells coordinate GPX4 synthesis with nutrient availability remains unclear. In this study, we perform integrated proteomic and functional analyses to reveal that SLC7A11-mediated cystine uptake promotes not only GSH synthesis, but also GPX4 protein synthesis. Mechanistically, we find that cyst(e)ine activates mechanistic/mammalian target of rapamycin complex 1 (mTORC1) and promotes GPX4 protein synthesis at least partly through the Rag-mTORC1-4EBP signaling axis. We show that pharmacologic inhibition of mTORC1 decreases GPX4 protein levels, sensitizes cancer cells to ferroptosis, and synergizes with ferroptosis inducers to suppress patient-derived xenograft tumor growth in vivo. Together, our results reveal a regulatory mechanism to coordinate GPX4 protein synthesis with cyst(e)ine availability and suggest using combinatorial therapy of mTORC1 inhibitors and ferroptosis inducers in cancer treatment.

Bipartite graph-based approach for clustering of cell lines by gene expression-drug response associations

MOTIVATION

In pharmacogenomic studies, the biological context of cell lines influences the predictive ability of drug-response models and the discovery of biomarkers. Thus, similar cell lines are often studied together based on prior knowledge of biological annotations. However, this selection approach is not scalable with the number of annotations, and the relationship between gene-drug association patterns and biological context may not be obvious.

RESULTS

We present a procedure to compare cell lines based on their gene-drug association patterns. Starting with a grouping of cell lines from biological annotation, we model gene-drug association patterns for each group as a bipartite graph between genes and drugs. This is accomplished by applying sparse canonical correlation analysis (SCCA) to extract the gene-drug associations, and using the canonical vectors to construct the edge weights. Then, we introduce a nuclear norm-based dissimilarity measure to compare the bipartite graphs. Accompanying our procedure is a permutation test to evaluate the significance of similarity of cell line groups in terms of gene-drug associations. In the pharmacogenomics datasets CTRP2, GDSC2, and CCLE, hierarchical clustering of carcinoma groups based on this dissimilarity measure uniquely reveals clustering patterns driven by carcinoma subtype rather than primary site. Next, we show that the top associated drugs or genes from SCCA can be used to characterize the clustering patterns of haematopoietic and lymphoid malignancies. Finally, we confirm by simulation that when drug responses are linearly-dependent on expression, our approach is the only one that can effectively infer the true hierarchy compared to existing approaches.

AVAILABILITY

Bipartite graph-based hierarchical clustering is implemented in R and can be obtained from CRAN: https://CRAN.R-project.org/package=hierBipartite. The source code is available at https://github.com/CalvinTChi/hierBipartite.

SUPPLEMENTARY INFORMATION

Supplementary data are available at Bioinformatics online.

Identification of tumor mutation burden-related hub genes and the underlying mechanism in melanoma

Background: Tumor mutation burden (TMB) has emerged as an important predictive factor for drug resistance in cancers; however, the specific mechanism underlying TMB function in melanoma remains elusive. Methods: Data on somatic mutations, RNA sequencing (RNA-seq), miRNA sequencing (miRNA-seq), and clinical characteristics for 472 melanoma patients were extracted from the TCGA cohort. RNA-seq data of melanoma cell lines were obtained from the Cancer Cell Line Encyclopedia, and sensitivity of cell lines to therapeutic agents is available in the Cancer Therapeutics Response Portal. TMB was calculated based on somatic mutation data. Differentially expressed gene analysis, weighted gene co-expression network analysis, protein-protein interaction networks, Minimal Common Oncology Data Elements, and survival analysis were leveraged to determine TMB-related hub genes. Competing endogenous RNA (ceRNA) networks were constructed to explore the molecular mechanisms underlying hub gene function. The influence of key genes on drug sensitivity was analyzed to investigate their clinical significance. Results: Elevated TMB levels were significantly correlated with improved survival outcomes. In addition, six tumor-infiltrating immune cells, including naive B cells, regulatory T cells, memory resting CD4 T cells, memory B cells, activated mast cells, and resting NK cells, were significantly overexpressed in the low-TMB group relative to the high-TMB group. Furthermore, we identified FLNC, NEXN, and TNNT3 as TMB-related hub genes, and constructed their ceRNA networks, including five miRNAs (has-miR-590-3p, has-miR-374b-5p, has-miR-3127-5p, has-miR-1913, and has-miR-1291) and 31 lncRNAs (FAM66C, MIAT, NR2F2AS1, etc.). Finally, we observed that TMB-related genes were associated with distinct therapeutic responses to AKT/mTOR pathway inhibitors. Conclusions: We identified three TMB-associated key genes, established their ceRNA networks, and investigated their influence on therapeutic responses, which could provide insights into future precision medicine.

Morphological profiling of small molecules

Profiling approaches such as gene expression or proteome profiling generate small-molecule bioactivity profiles that describe a perturbed cellular state in a rather unbiased manner and have become indispensable tools in the evaluation of bioactive small molecules. Automated imaging and image analysis can record morphological alterations that are induced by small molecules through the detection of hundreds of morphological features in high-throughput experiments. Thus, morphological profiling has gained growing attention in academia and the pharmaceutical industry as it enables detection of bioactivity in compound collections in a broader biological context in the early stages of compound development and the drug-discovery process. Profiling may be used successfully to predict mode of action or targets of unexplored compounds and to uncover unanticipated activity for already characterized small molecules. Here, we review the reported approaches to morphological profiling and the kind of bioactivity that can be detected so far and, thus, predicted.

Multi-modal meta-analysis of cancer cell line omics profiles identifies ECHDC1 as a novel breast tumor suppressor

Molecular and functional profiling of cancer cell lines is subject to laboratory-specific experimental practices and data analysis protocols. The current challenge therefore is how to make an integrated use of the omics profiles of cancer cell lines for reliable biological discoveries. Here, we carried out a systematic analysis of nine types of data modalities using meta-analysis of 53 omics studies across 12 research laboratories for 2,018 cell lines. To account for a relatively low consistency observed for certain data modalities, we developed a robust data integration approach that identifies reproducible signals shared among multiple data modalities and studies. We demonstrated the power of the integrative analyses by identifying a novel driver gene, ECHDC1, with tumor suppressive role validated both in breast cancer cells and patient tumors. The multi-modal meta-analysis approach also identified synthetic lethal partners of cancer drivers, including a co-dependency of PTEN deficient endometrial cancer cells on RNA helicases.

Facilitating Drug Discovery in Breast Cancer by Virtually Screening Patients Using In Vitro Drug Response Modeling

(1) Background: Drug imputation methods often aim to translate in vitro drug response to in vivo drug efficacy predictions. While commonly used in retrospective analyses, our aim is to investigate the use of drug prediction methods for the generation of novel drug discovery hypotheses. Triple-negative breast cancer (TNBC) is a severe clinical challenge in need of new therapies. (2) Methods: We used an established machine learning approach to build models of drug response based on cell line transcriptome data, which we then applied to patient tumor data to obtain predicted sensitivity scores for hundreds of drugs in over 1000 breast cancer patients. We then examined the relationships between predicted drug response and patient clinical features. (3) Results: Our analysis recapitulated several suspected vulnerabilities in TNBC and identified a number of compounds-of-interest. AZD-1775, a Wee1 inhibitor, was predicted to have preferential activity in TNBC (p < 2.2 × 10-16) and its efficacy was highly associated with TP53 mutations (p = 1.2 × 10-46). We validated these findings using independent cell line screening data and pathway analysis. Additionally, co-administration of AZD-1775 with standard-of-care paclitaxel was able to inhibit tumor growth (p < 0.05) and increase survival (p < 0.01) in a xenograft mouse model of TNBC. (4) Conclusions: Overall, this study provides a framework to turn any cancer transcriptomic dataset into a dataset for drug discovery. Using this framework, one can quickly generate meaningful drug discovery hypotheses for a cancer population of interest.

Towards the routine use of in silico screenings for drug discovery using metabolic modelling

Currently, the development of new effective drugs for cancer therapy is not only hindered by development costs, drug efficacy, and drug safety but also by the rapid occurrence of drug resistance in cancer. Hence, new tools are needed to study the underlying mechanisms in cancer. Here, we discuss the current use of metabolic modelling approaches to identify cancer-specific metabolism and find possible new drug targets and drugs for repurposing. Furthermore, we list valuable resources that are needed for the reconstruction of cancer-specific models by integrating various available datasets with genome-scale metabolic reconstructions using model-building algorithms. We also discuss how new drug targets can be determined by using gene essentiality analysis, an in silico method to predict essential genes in a given condition such as cancer and how synthetic lethality studies could greatly benefit cancer patients by suggesting drug combinations with reduced side effects.

Biological and therapeutic implications of a unique subtype of NPM1 mutated AML

In acute myeloid leukemia (AML), molecular heterogeneity across patients constitutes a major challenge for prognosis and therapy. AML with NPM1 mutation is a distinct genetic entity in the revised World Health Organization classification. However, differing patterns of co-mutation and response to therapy within this group necessitate further stratification. Here we report two distinct subtypes within NPM1 mutated AML patients, which we label as primitive and committed based on the respective presence or absence of a stem cell signature. Using gene expression (RNA-seq), epigenomic (ATAC-seq) and immunophenotyping (CyToF) analysis, we associate each subtype with specific molecular characteristics, disease differentiation state and patient survival. Using ex vivo drug sensitivity profiling, we show a differential drug response of the subtypes to specific kinase inhibitors, irrespective of the FLT3-ITD status. Differential drug responses of the primitive and committed subtype are validated in an independent AML cohort. Our results highlight heterogeneity among NPM1 mutated AML patient samples based on stemness and suggest that the addition of kinase inhibitors to the treatment of cases with the primitive signature, lacking FLT3-ITD, could have therapeutic benefit.

Artificial intelligence, machine learning, and drug repurposing in cancer

Introduction: Drug repurposing provides a cost-effective strategy to re-use approved drugs for new medical indications. Several machine learning (ML) and artificial intelligence (AI) approaches have been developed for systematic identification of drug repurposing leads based on big data resources, hence further accelerating and de-risking the drug development process by computational means.Areas covered: The authors focus on supervised ML and AI methods that make use of publicly available databases and information resources. While most of the example applications are in the field of anticancer drug therapies, the methods and resources reviewed are widely applicable also to other indications including COVID-19 treatment. A particular emphasis is placed on the use of comprehensive target activity profiles that enable a systematic repurposing process by extending the target profile of drugs to include potent off-targets with therapeutic potential for a new indication.Expert opinion: The scarcity of clinical patient data and the current focus on genetic aberrations as primary drug targets may limit the performance of anticancer drug repurposing approaches that rely solely on genomics-based information. Functional testing of cancer patient cells exposed to a large number of targeted therapies and their combinations provides an additional source of repurposing information for tissue-aware AI approaches.

Current status and future outlook for patient-derived cancer models from a rare cancer research perspective

Rare cancers are a group of approximately 200 malignancies with extremely low incidences and with a wide variety of genotypes and phenotypes. Collectively, they are more common than any single malignancy. However, given the small numbers of individuals diagnosed with rare cancers, it is difficult to design clinical trials with sufficient patient numbers. Therefore, few effective anticancer drugs have been developed, and evidence‐based medicine is not always feasible for rare cancers. Consequently, their clinical outcomes are generally poorer. Cancer research requires adequate models that faithfully recapitulate molecular features and reproduce treatment responses of the original tumors. Such models allow us to focus on more efficacious drugs in the clinical studies. For rare cancers, patient‐derived cancer models are particularly important because the enrollment of sufficient patients is rarely attainable within a reasonable period of time. However, extremely few models are available for rare cancers. For example, cell lines and xenografts are available for only a limited number of histological subtypes of sarcomas; therefore, most sarcoma research is performed without such models, and a lack of adequate cancer models causes a lag in therapeutic development. The establishment of novel rare cancer models will dramatically facilitate rare cancer research and treatment development in the near future. This review focuses on the status of patient‐derived rare cancer models and discusses their pivotal problems and possibilities, using sarcomas as a representative rare cancer type. Multi‐institutional collaboration will help address the scarcity of patient‐derived rare cancer models.

A thiol-bound drug reservoir enhances APR-246-induced mutant p53 tumor cell death

The tumor suppressor gene TP53 is the most frequently mutated gene in cancer. The compound APR-246 (PRIMA-1Met/Eprenetapopt) is converted to methylene quinuclidinone (MQ) that targets mutant p53 protein and perturbs cellular antioxidant balance. APR-246 is currently tested in a phase III clinical trial in myelodysplastic syndrome (MDS). By in vitro, ex vivo, and in vivo models, we show that combined treatment with APR-246 and inhibitors of efflux pump MRP1/ABCC1 results in synergistic tumor cell death, which is more pronounced in TP53 mutant cells. This is associated with altered cellular thiol status and increased intracellular glutathione-conjugated MQ (GS-MQ). Due to the reversibility of MQ conjugation, GS-MQ forms an intracellular drug reservoir that increases availability of MQ for targeting mutant p53. Our study shows that redox homeostasis is a critical determinant of the response to mutant p53-targeted cancer therapy.

LincRNA-immunity landscape analysis identifies EPIC1 as a regulator of tumor immune evasion and immunotherapy resistance

Through an integrative analysis of the lincRNA expression and tumor immune response in 9,626 tumor samples across 32 cancer types, we developed a lincRNA-based immune response (LIMER) score that can predict the immune cells infiltration and patient prognosis in multiple cancer types. Our analysis also identified tumor-specific lincRNAs, including EPIC1, that potentially regulate tumor immune response in multiple cancer types. Immunocompetent mouse models and in vitro co-culture assays demonstrated that EPIC1 induces tumor immune evasion and resistance to immunotherapy by suppressing tumor cell antigen presentation. Mechanistically, lincRNA EPIC1 interacts with the histone methyltransferase EZH2, leading to the epigenetic silencing of IFNGR1, TAP1/2, ERAP1/2, and MHC-I genes. Genetic and pharmacological inhibition of EZH2 abolish EPIC1's immune-related oncogenic effect and its suppression of interferon-γ signaling. The EPIC1-EZH2 axis emerges as a potential mechanism for tumor immune evasion that can serve as therapeutic targets for immunotherapy.

Cross-species identification of PIP5K1-, splicing- and ubiquitin-related pathways as potential targets for RB1-deficient cells

The RB1 tumor suppressor is recurrently mutated in a variety of cancers including retinoblastomas, small cell lung cancers, triple-negative breast cancers, prostate cancers, and osteosarcomas. Finding new synthetic lethal (SL) interactions with RB1 could lead to new approaches to treating cancers with inactivated RB1. We identified 95 SL partners of RB1 based on a Drosophila screen for genetic modifiers of the eye phenotype caused by defects in the RB1 ortholog, Rbf1. We validated 38 mammalian orthologs of Rbf1 modifiers as RB1 SL partners in human cancer cell lines with defective RB1 alleles. We further show that for many of the RB1 SL genes validated in human cancer cell lines, low activity of the SL gene in human tumors, when concurrent with low levels of RB1 was associated with improved patient survival. We investigated higher order combinatorial gene interactions by creating a novel Drosophila cancer model with co-occurring Rbf1, Pten and Ras mutations, and found that targeting RB1 SL genes in this background suppressed the dramatic tumor growth and rescued fly survival whilst having minimal effects on wild-type cells. Finally, we found that drugs targeting the identified RB1 interacting genes/pathways, such as UNC3230, PYR-41, TAK-243, isoginkgetin, madrasin, and celastrol also elicit SL in human cancer cell lines. In summary, we identified several high confidence, evolutionarily conserved, novel targets for RB1-deficient cells that may be further adapted for the treatment of human cancer.

The RNA m6A Reader YTHDF2 Maintains Oncogene Expression and Is a Targetable Dependency in Glioblastoma Stem Cells

Glioblastoma is a universally lethal cancer driven by glioblastoma stem cells (GSC). Here, we interrogated N 6-methyladenosine (m6A) mRNA modifications in GSCs by methyl RNA immunoprecipitation followed by sequencing and transcriptome analysis, finding transcripts marked by m6A often upregulated compared with normal neural stem cells (NSC). Interrogating m6A regulators, GSCs displayed preferential expression, as well as in vitro and in vivo dependency, of the m6A reader YTHDF2, in contrast to NSCs. Although YTHDF2 has been reported to destabilize mRNAs, YTHDF2 stabilized MYC and VEGFA transcripts in GSCs in an m6A-dependent manner. We identified IGFBP3 as a downstream effector of the YTHDF2-MYC axis in GSCs. The IGF1/IGF1R inhibitor linsitinib preferentially targeted YTHDF2-expressing cells, inhibiting GSC viability without affecting NSCs and impairing in vivo glioblastoma growth. Thus, YTHDF2 links RNA epitranscriptomic modifications and GSC growth, laying the foundation for the YTHDF2-MYC-IGFBP3 axis as a specific and novel therapeutic target in glioblastoma. SIGNIFICANCE: Epitranscriptomics promotes cellular heterogeneity in cancer. RNA m6A landscapes of cancer and NSCs identified cell type-specific dependencies and therapeutic vulnerabilities. The m6A reader YTHDF2 stabilized MYC mRNA specifically in cancer stem cells. Given the challenge of targeting MYC, YTHDF2 presents a therapeutic target to perturb MYC signaling in glioblastoma.This article is highlighted in the In This Issue feature, p. 211.

Aneuploidy renders cancer cells vulnerable to mitotic checkpoint inhibition

Selective targeting of aneuploid cells is an attractive strategy for cancer treatment1. However, it is unclear whether aneuploidy generates any clinically relevant vulnerabilities in cancer cells. Here we mapped the aneuploidy landscapes of about 1,000 human cancer cell lines, and analysed genetic and chemical perturbation screens2-9 to identify cellular vulnerabilities associated with aneuploidy. We found that aneuploid cancer cells show increased sensitivity to genetic perturbation of core components of the spindle assembly checkpoint (SAC), which ensures the proper segregation of chromosomes during mitosis10. Unexpectedly, we also found that aneuploid cancer cells were less sensitive than diploid cells to short-term exposure to multiple SAC inhibitors. Indeed, aneuploid cancer cells became increasingly sensitive to inhibition of SAC over time. Aneuploid cells exhibited aberrant spindle geometry and dynamics, and kept dividing when the SAC was inhibited, resulting in the accumulation of mitotic defects, and in unstable and less-fit karyotypes. Therefore, although aneuploid cancer cells could overcome inhibition of SAC more readily than diploid cells, their long-term proliferation was jeopardized. We identified a specific mitotic kinesin, KIF18A, whose activity was perturbed in aneuploid cancer cells. Aneuploid cancer cells were particularly vulnerable to depletion of KIF18A, and KIF18A overexpression restored their response to SAC inhibition. Our results identify a therapeutically relevant, synthetic lethal interaction between aneuploidy and the SAC.

Drug-induced cell viability prediction from LINCS-L1000 through WRFEN-XGBoost algorithm

Background

Predicting the drug response of the cancer diseases through the cellular perturbation signatures under the action of specific compounds is very important in personalized medicine. In the process of testing drug responses to the cancer, traditional experimental methods have been greatly hampered by the cost and sample size. At present, the public availability of large amounts of gene expression data makes it a challenging task to use machine learning methods to predict the drug sensitivity.

Results

In this study, we introduced the WRFEN-XGBoost cell viability prediction algorithm based on LINCS-L1000 cell signatures. We integrated the LINCS-L1000, CTRP and Achilles datasets and adopted a weighted fusion algorithm based on random forest and elastic net for key gene selection. Then the FEBPSO algorithm was introduced into XGBoost learning algorithm to predict the cell viability induced by the drugs. The proposed method was compared with some new methods, and it was found that our model achieved good results with 0.83 Pearson correlation. At the same time, we completed the drug sensitivity validation on the NCI60 and CCLE datasets, which further demonstrated the effectiveness of our method.

Conclusions

The results showed that our method was conducive to the elucidation of disease mechanisms and the exploration of new therapies, which greatly promoted the progress of clinical medicine.

OCTAD: an open workspace for virtually screening therapeutics targeting precise cancer patient groups using gene expression features

As the field of precision medicine progresses, treatments for patients with cancer are starting to be tailored to their molecular as well as their clinical features. The emerging cancer subtypes defined by these molecular features require that dedicated resources be used to assist the discovery of drug candidates for preclinical evaluation. Voluminous gene expression profiles of patients with cancer have been accumulated in public databases, enabling the creation of cancer-specific expression signatures. Meanwhile, large-scale gene expression profiles of cellular responses to chemical compounds have also recently became available. By matching the cancer-specific expression signature to compound-induced gene expression profiles from large drug libraries, researchers can prioritize small molecules that present high potency to reverse expression of signature genes for further experimental testing of their efficacy. This approach has proven to be an efficient and cost-effective way to identify efficacious drug candidates. However, the success of this approach requires multiscale procedures, imposing considerable challenges to many labs. To address this, we developed Open Cancer TherApeutic Discovery (OCTAD; http://octad.org ): an open workspace for virtually screening compounds targeting precise groups of patients with cancer using gene expression features. Its database includes 19,127 patient tissue samples covering more than 50 cancer types and expression profiles for 12,442 distinct compounds. The program is used to perform deep-learning-based reference tissue selection, disease gene expression signature creation, drug reversal potency scoring and in silico validation. OCTAD is available as a web portal and a standalone R package to allow experimental and computational scientists to easily navigate the tool.

Identification of Predictive ERBB Mutations by Leveraging Publicly Available Cell Line Databases

Although targeted therapies can be effective for a subgroup of patients, identification of individuals who benefit from the treatments is challenging. At the same time, the predictive significance of the majority of the thousands of mutations observed in the cancer tissues remains unknown. Here, we describe the identification of novel predictive biomarkers for ERBB-targeted tyrosine kinase inhibitors (TKIs) by leveraging the genetic and drug screening data available in the public cell line databases: Cancer Cell Line Encyclopedia, Genomics of Drug Sensitivity in Cancer, and Cancer Therapeutics Response Portal. We assessed the potential of 412 ERBB mutations in 296 cell lines to predict responses to 10 different ERBB-targeted TKIs. Seventy-six ERBB mutations were identified that were associated with ERBB TKI sensitivity comparable with non-small cell lung cancer cell lines harboring the well-established predictive EGFR L858R mutation or exon 19 deletions. Fourteen (18.4%) of these mutations were classified as oncogenic by the cBioPortal database, whereas 62 (81.6%) were regarded as novel potentially predictive mutations. Of the nine functionally validated novel mutations, EGFR Y1069C and ERBB2 E936K were transforming in Ba/F3 cells and demonstrated enhanced signaling activity. Mechanistically, the EGFR Y1069C mutation disrupted the binding of the ubiquitin ligase c-CBL to EGFR, whereas the ERBB2 E936K mutation selectively enhanced the activity of ERBB heterodimers. These findings indicate that integrating data from publicly available cell line databases can be used to identify novel, predictive nonhotspot mutations, potentially expanding the patient population benefiting from existing cancer therapies.

CDKN2A-Inactivated Pancreatic Ductal Adenocarcinoma Exhibits Therapeutic Sensitivity to Paclitaxel: A Bioinformatics Study

The mutation of cyclin dependent kinase inhibitor 2A (CDKN2A) is frequently found in pancreatic ductal adenocarcinoma (PDAC). However, its prognostic and therapeutic roles in PDAC have not been extensively investigated yet. In this study, we mined and integrated the cancer genomics and chemogenomics data to investigate the roles of CDKN2A genetic alterations in PDAC patients' prognosis and treatment. We found that functional CDKN2A inactivation caused by mutations and deep deletions predicted poor prognosis in PDAC patients. CDKN2A inactivation was associated with the upregulation of genes related to estrogen response, which can be overcome by CDKN2A restoration. Chemosensitivity profiling of PDAC cell lines and patient-derived organoids found that CDKN2A inactivation was associated with the increased sensitivity to paclitaxel and SN-38 (the active metabolite of irinotecan). However, only paclitaxel can mimic the effect of CDKN2A restoration, and its drug sensitivity was correlated with genes related to estrogen response. Therefore, our study suggested that CDKN2A-inactivated PDAC patients could benefit from the precision treatment with paclitaxel, whose albumin-stabilized nanoparticle formulation (nab-paclitaxel) has been approved for treating PDAC.

A statistical framework for assessing pharmacological responses and biomarkers using uncertainty estimates

High-throughput testing of drugs across molecular-characterised cell lines can identify candidate treatments and discover biomarkers. However, the cells' response to a drug is typically quantified by a summary statistic from a best-fit dose-response curve, whilst neglecting the uncertainty of the curve fit and the potential variability in the raw readouts. Here, we model the experimental variance using Gaussian Processes, and subsequently, leverage uncertainty estimates to identify associated biomarkers with a new Bayesian framework. Applied to in vitro screening data on 265 compounds across 1074 cancer cell lines, our models identified 24 clinically established drug-response biomarkers, and provided evidence for six novel biomarkers by accounting for association with low uncertainty. We validated our uncertainty estimates with an additional drug screen of 26 drugs, 10 cell lines with 8 to 9 replicates. Our method is applicable to any dose-response data without replicates, and improves biomarker discovery for precision medicine.

Deep learning of pharmacogenomics resources: moving towards precision oncology

The recent accumulation of cancer genomic data provides an opportunity to understand how a tumor's genomic characteristics can affect its responses to drugs. This field, called pharmacogenomics, is a key area in the development of precision oncology. Deep learning (DL) methodology has emerged as a powerful technique to characterize and learn from rapidly accumulating pharmacogenomics data. We introduce the fundamentals and typical model architectures of DL. We review the use of DL in classification of cancers and cancer subtypes (diagnosis and treatment stratification of patients), prediction of drug response and drug synergy for individual tumors (treatment prioritization for a patient), drug repositioning and discovery and the study of mechanism/mode of action of treatments. For each topic, we summarize current genomics and pharmacogenomics data resources such as pan-cancer genomics data for cancer cell lines (CCLs) and tumors, and systematic pharmacologic screens of CCLs. By revisiting the published literature, including our in-house analyses, we demonstrate the unprecedented capability of DL enabled by rapid accumulation of data resources to decipher complex drug response patterns, thus potentially improving cancer medicine. Overall, this review provides an in-depth summary of state-of-the-art DL methods and up-to-date pharmacogenomics resources and future opportunities and challenges to realize the goal of precision oncology.