More than fishing for a cure: The promises and pitfalls of high throughput cancer cell line screens

Computational drug discovery pipelines identify NAMPT as a therapeutic target in neuroendocrine prostate cancer

Neuroendocrine prostate cancer (NEPC) is an aggressive advanced subtype of prostate cancer that exhibits poor prognosis and broad resistance to therapies. Currently, few treatment options are available, highlighting a need for new therapeutics to help curb the high mortality rates of this disease. We designed a comprehensive drug discovery pipeline that quickly generates drug candidates ready to be tested. Our method estimated patient response to various therapeutics in three independent prostate cancer patient cohorts and selected robust candidate drugs showing high predicted potency in NEPC tumors. Using this pipeline, we nominated NAMPT as a molecular target to effectively treat NEPC tumors. Our in vitro experiments validated the efficacy of NAMPT inhibitors in NEPC cells. Compared with adenocarcinoma LNCaP cells, NAMPT inhibitors induced significantly higher growth inhibition in the NEPC cell line model NCI-H660. Moreover, to further assist clinical development, we implemented a causal feature selection method to detect biomarkers indicative of sensitivity to NAMPT inhibitors. Gene expression modifications of selected biomarkers resulted in changes in sensitivity to NAMPT inhibitors consistent with expectations in NEPC cells. Validation of these markers in an independent prostate cancer patient dataset supported their use to inform clinical efficacy. Our findings pave the way for new treatments to combat pervasive drug resistance and reduce mortality. Furthermore, this research highlights the use of drug sensitivity-related biomarkers to understand mechanisms and potentially indicate clinical efficacy.

Bayesian inference for multivariate probit model with latent envelope

The response envelope model proposed by Cook et al. (2010) is an efficient method to estimate the regression coefficient under the context of the multivariate linear regression model. It improves estimation efficiency by identifying material and immaterial parts of responses and removing the immaterial variation. The response envelope model has been investigated only for continuous response variables. In this paper, we propose the multivariate probit model with latent envelope, in short, the probit envelope model, as a response envelope model for multivariate binary response variables. The probit envelope model takes into account relations between Gaussian latent variables of the multivariate probit model by using the idea of the response envelope model. We address the identifiability of the probit envelope model by employing the essential identifiability concept and suggest a Bayesian method for the parameter estimation. We illustrate the probit envelope model via simulation studies and real-data analysis. The simulation studies show that the probit envelope model has the potential to gain efficiency in estimation compared to the multivariate probit model. The real data analysis shows that the probit envelope model is useful for multi-label classification.

CODEX: COunterfactual Deep learning for the in silico EXploration of cancer cell line perturbations

MOTIVATION

High-throughput screens (HTS) provide a powerful tool to decipher the causal effects of chemical and genetic perturbations on cancer cell lines. Their ability to evaluate a wide spectrum of interventions, from single drugs to intricate drug combinations and CRISPR-interference, has established them as an invaluable resource for the development of novel therapeutic approaches. Nevertheless, the combinatorial complexity of potential interventions makes a comprehensive exploration intractable. Hence, prioritizing interventions for further experimental investigation becomes of utmost importance.

RESULTS

We propose CODEX (COunterfactual Deep learning for the in silico EXploration of cancer cell line perturbations) as a general framework for the causal modeling of HTS data, linking perturbations to their downstream consequences. CODEX relies on a stringent causal modeling strategy based on counterfactual reasoning. As such, CODEX predicts drug-specific cellular responses, comprising cell survival and molecular alterations, and facilitates the in silico exploration of drug combinations. This is achieved for both bulk and single-cell HTS. We further show that CODEX provides a rationale to explore complex genetic modifications from CRISPR-interference in silico in single cells.

AVAILABILITY AND IMPLEMENTATION

Our implementation of CODEX is publicly available at https://github.com/sschrod/CODEX. All data used in this article are publicly available.

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.

Identification of hub genes related to metastasis and prognosis of osteosarcoma and establishment of a prognostic model with bioinformatic methods

Osteosarcoma (OS) is the most common primary malignant bone tumor occurring in children and adolescents. Improvements in our understanding of the OS pathogenesis and metastatic mechanism on the molecular level might lead to notable advances in the treatment and prognosis of OS. Biomarkers related to OS metastasis and prognosis were analyzed and identified, and a prognostic model was established through the integration of bioinformatics tools and datasets in multiple databases. 2 OS datasets were downloaded from the Gene Expression Omnibus database for data consolidation, standardization, batch effect correction, and identification of differentially expressed genes (DEGs); following that, gene ontology and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway enrichment analyses were performed on the DEGs; the STRING database was subsequently used for protein-protein interaction (PPI) network construction and identification of hub genes; hub gene expression was validated, and survival analysis was conducted through the employment of the TARGET database; finally, a prognostic model was established and evaluated subsequent to the screening of survival-related genes. A total of 701 DEGs were identified; by gene ontology and KEGG pathway enrichment analyses, the overlapping DEGs were enriched for 249 biological process terms, 13 cellular component terms, 35 molecular function terms, and 4 KEGG pathways; 13 hub genes were selected from the PPI network; 6 survival-related genes were identified by the survival analysis; the prognostic model suggested that 4 genes were strongly associated with the prognosis of OS. DEGs related to OS metastasis and survival were identified through bioinformatics analysis, and hub genes were further selected to establish an ideal prognostic model for OS patients. On this basis, 4 protective genes including TPM1, TPM2, TPM3, and TPM4 were yielded by the prognostic model.

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.

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.

Multi-omics analyses of choroid plexus carcinoma cell lines reveal potential targetable pathways and alterations

PURPOSE

Choroid plexus carcinomas (CPCs) are extremely rare brain tumors and carry a dismal prognosis. Treatment options are limited and there is an urgent need to develop models to further research. In the present study, we established two CPC cell lines and performed multi-omics analyses. These cell lines serve as valuable models to propose new treatments in these rare but deadly brain tumors.

METHODS

Multi-omic profiling including, (i) methylation array (EPIC 850 K), (ii) whole genome sequencing (WGS), (iii) CANCERPLEX cancer genome panel testing, (iv) RNA sequencing (RNA-seq), and (v) proteomics analyses were performed in CCHE-45 and NGT131 cell lines.

RESULTS

Both cell lines were classified as methylation class B. Both harbored pathogenic TP53 point mutations; CCHE-45 additionally displayed TP53 loss. Furthermore, alterations of the NOTCH and WNT pathways were also detected in both cell lines. Two protein-coding gene fusions, BZW2-URGCP, and CTTNBP2-ERBB4, mutations of two oncodrivers, GBP-4 and KRTAP-12-2, and several copy number alterations were observed in CCHE-45, but not NGT131. Transcriptome and proteome analysis identified shared and unique signatures, suggesting that variability in choroid plexus carcinoma tumors may exist. The discovered difference's importance and implications highlight the possible diversity of choroid plexus carcinoma and call for additional research to fully understand disease pathogenesis.

CONCLUSION

Multi-omics analyses revealed that the two choroid plexus carcinoma cell lines shared TP53 mutations and other common pathway alterations and activation of NOTCH and WNT pathways. Noticeable differences were also observed. These cell lines can serve as valuable models to propose new treatments in these rare but deadly brain tumors.

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/.

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

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

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

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

Multiomic Investigations into Lung Health and Disease

Diseases of the lung account for more than 5 million deaths worldwide and are a healthcare burden. Improving clinical outcomes, including mortality and quality of life, involves a holistic understanding of the disease, which can be provided by the integration of lung multi-omics data. An enhanced understanding of comprehensive multiomic datasets provides opportunities to leverage those datasets to inform the treatment and prevention of lung diseases by classifying severity, prognostication, and discovery of biomarkers. The main objective of this review is to summarize the use of multiomics investigations in lung disease, including multiomics integration and the use of machine learning computational methods. This review also discusses lung disease models, including animal models, organoids, and single-cell lines, to study multiomics in lung health and disease. We provide examples of lung diseases where multi-omics investigations have provided deeper insight into etiopathogenesis and have resulted in improved preventative and therapeutic interventions.

A Comprehensive Review of Interventional Clinical Trials in Patients with Bone Metastases

Bone metastasis is one of the most important factors associated with poor prognosis for patients with prostate, breast, thyroid, and lung cancer. In the past two decades, 651 clinical trials, including 554 interventional trials, were being registered in ClinicalTrials.gov and pharma.id.informa.com to combat bone metastases from different perspectives. In this review, we comprehensively analyzed, regrouped, and discussed all the interventional trials on bone metastases. Clinical trials were re-grouped into bone-targeting agents, radiotherapy, small molecule targeted therapy, combination therapy, and others, based on the different mechanisms of action including modifying the bone microenvironment and preventing the growth of cancer cells. We also discussed the potential strategies that might improve overall survival and progression-free survival of patients with bone metastases in the future.

Drosophila as a diet discovery tool for treating amino acid disorders

Amino acid disorders (AADs) are a large group of rare inherited conditions that collectively impact one in 6500 live births, often resulting in rapid neurological decline and death during infancy. For several AADs, including phenylketonuria, dietary modification prevents physiological deterioration and ameliorates symptoms. Despite this remarkable potential for treatment success, dietary therapy for most AADs remains largely unexplored. Although animal models have provided novel insights into AAD mechanisms, few have been used for therapeutic diet discovery. Here, we find that of all the animal models, Drosophila is particularly well suited for nutrigenomic disease modelling, having amino acid pathways conserved with humans, exceptional genetic tractability, and the unique availability of a synthetic customisable diet.

High-throughput screen in vitro identifies dasatinib as a candidate for combinatorial treatment with HER2-targeting drugs in breast cancer

Human epidermal growth factor receptor 2-positive (HER2+) breast cancer is an aggressive subtype of this disease. Targeted treatment has improved outcome, but there is still a need for new therapeutic strategies as some patients respond poorly to treatment. Our aim was to identify compounds that substantially affect viability in HER2+ breast cancer cells in response to combinatorial treatment. We performed a high-throughput drug screen of 278 compounds in combination with trastuzumab and lapatinib using two HER2+ breast cancer cell lines (KPL4 and SUM190PT). The most promising drugs were validated in vitro and in vivo, and downstream molecular changes of the treatments were analyzed. The screen revealed multiple drugs that could be used in combination with lapatinib and/or trastuzumab. The Src-inhibitor dasatinib showed the largest combinatorial effect together with lapatinib in the KPL4 cell line compared to treatment with dasatinib alone (p < 0.01). In vivo, only lapatinib significantly reduced tumor growth (p < 0.05), whereas dasatinib alone, or in combination with lapatinib, did not show significant effects. Protein analyses of the treated xenografts showed significant alterations in protein levels compared to untreated controls, suggesting that all drugs reached the tumor and exerted a measurable effect. In silico analyses suggested activation of apoptosis and reduced activity of survival pathways by all treatments, but the opposite pattern was observed for the combinatorial treatment compared to lapatinib alone.

Drug screening and genome editing in human pancreatic cancer organoids identifies drug-gene interactions and candidates for off-label treatment

Pancreatic cancer (PDAC) is a highly aggressive malignancy for which the identification of novel therapies is urgently needed. Here, we establish a human PDAC organoid biobank from 31 genetically distinct lines, covering a representative range of tumor subtypes, and demonstrate that these reflect the molecular and phenotypic heterogeneity of primary PDAC tissue. We use CRISPR-Cas9 genome editing and drug screening to characterize drug-gene interactions with ARID1A and BRCA2. We find that missense- but not frameshift mutations in the PDAC driver gene ARID1A are associated with increased sensitivity to the kinase inhibitors dasatinib (p < 0.0001) and VE-821 (p < 0.0001). We conduct an automated drug-repurposing screen with 1,172 FDA-approved compounds, identifying 26 compounds that effectively kill PDAC organoids, including 19 chemotherapy drugs currently approved for other cancer types. We validate the activity of these compounds in vitro and in vivo. The in vivo validated hits include emetine and ouabain, compounds which are approved for non-cancer indications and which perturb the ability of PDAC organoids to respond to hypoxia. Our study provides proof-of-concept for advancing precision oncology and identifying candidates for drug repurposing via genome editing and drug screening in tumor organoid biobanks.

Identification of hub genes for early detection of bone metastasis in breast cancer

BACKGROUND

Globally, among all women, the most frequently detected and diagnosed and the most lethal type of cancer is breast cancer (BC). In particular, bone is one of the most frequent distant metastases 24in breast cancer patients and bone metastasis arises in approximately 80% of advanced patients. Thus, we need to identify and validate early detection markers that can differentiate metastasis from non-metastasis breast cancers.

METHODS

GSE55715, GSE103357, and GSE146661 gene expression profiling data were downloaded from the GEO database. There was 14 breast cancer with bone metastasis samples and 8 breast cancer tissue samples. GEO2R was used to screen for differentially expressed genes (DEGs). The volcano plots, Venn diagrams, and annular heatmap were generated by using the ggplot2 package. By using the cluster Profiler R package, KEGG and GO enrichment analyses of DEGs were conducted. Through PPI network construction using the STRING database, key hub genes were identified by cytoHubba. Finally, K-M survival and ROC curves were generated to validate hub gene expression.

RESULTS

By GO enrichment analysis, 143 DEGs were enriched in the following GO terms: extracellular structure organization, extracellular matrix organization, leukocyte migration class II protein complex, collagen tridermic protein complex, extracellular matrix structural constituent, growth factor binding, and platelet-derived growth factor binding. In the KEGG pathway enrichment analysis, DEGs were enriched in Staphylococcus aureus infection, Complement and coagulation cascades, and Asthma. By PPI network analysis, we selected the top 10 genes, including SLCO2B1, STAB1, SERPING1, HLA-DOA, AIF1, GIMAP4, C1orf162, HLA-DMB, ADAP2, and HAVCR2. By using TCGA and THPA databases, we validated 2 genes, SERPING1 and GIMAP4, that were related to the early detection of bone metastasis in BC.

CONCLUSIONS

2 abnormally expressed hub genes could play a pivotal role in the breast cancer with bone metastasis by affecting bone homeostasis imbalance in the bone microenvironment.

Harnessing the predictive power of preclinical models for oncology drug development

Recent progress in understanding the molecular basis of cellular processes, identification of promising therapeutic targets and evolution of the regulatory landscape makes this an exciting and unprecedented time to be in the field of oncology drug development. However, high costs, long development timelines and steep rates of attrition continue to afflict the drug development process. Lack of predictive preclinical models is considered one of the key reasons for the high rate of attrition in oncology. Generating meaningful and predictive results preclinically requires a firm grasp of the relevant biological questions and alignment of the model systems that mirror the patient context. In doing so, the ability to conduct both forward translation, the process of implementing basic research discoveries into practice, as well as reverse translation, the process of elucidating the mechanistic basis of clinical observations, greatly enhances our ability to develop effective anticancer treatments. In this Review, we outline issues in preclinical-to-clinical translatability of molecularly targeted cancer therapies, present concepts and examples of successful reverse translation, and highlight the need to better align tumour biology in patients with preclinical model systems including tracking of strengths and weaknesses of preclinical models throughout programme development.

Identification of 5 Hub Genes Related to the Early Diagnosis, Tumour Stage, and Poor Outcomes of Hepatitis B Virus-Related Hepatocellular Carcinoma by Bioinformatics Analysis

BACKGROUND

The majority of primary liver cancers in adults worldwide are hepatocellular carcinomas (HCCs, or hepatomas). Thus, a deep understanding of the underlying mechanisms for the pathogenesis and carcinogenesis of HCC at the molecular level could facilitate the development of novel early diagnostic and therapeutic treatments to improve the approaches and prognosis for HCC patients. Our study elucidates the underlying molecular mechanisms of HBV-HCC development and progression and identifies important genes related to the early diagnosis, tumour stage, and poor outcomes of HCC.

METHODS

GSE55092 and GSE121248 gene expression profiling data were downloaded from the Gene Expression Omnibus (GEO) database. There were 119 HCC samples and 128 nontumour tissue samples. GEO2R was used to screen for differentially expressed genes (DEGs). Volcano plots and Venn diagrams were drawn by using the ggplot2 package in R. A heat map was generated by using Heatmapper. By using the clusterProfiler R package, KEGG and GO enrichment analyses of DEGs were conducted. Through PPI network construction using the STRING database, key hub genes were identified by cytoHubba. Finally, KM survival curves and ROC curves were generated to validate hub gene expression.

RESULTS

By GO enrichment analysis, 694 DEGs were enriched in the following GO terms: organic acid catabolic process, carboxylic acid catabolic process, carboxylic acid biosynthetic process, collagen-containing extracellular matrix, blood microparticle, condensed chromosome kinetochore, arachidonic acid epoxygenase activity, arachidonic acid monooxygenase activity, and monooxygenase activity. In the KEGG pathway enrichment analysis, DEGs were enriched in arachidonic acid epoxygenase activity, arachidonic acid monooxygenase activity, and monooxygenase activity. By PPI network construction and analysis of hub genes, we selected the top 10 genes, including CDK1, CCNB2, CDC20, BUB1, BUB1B, CCNB1, NDC80, CENPF, MAD2L1, and NUF2. By using TCGA and THPA databases, we found five genes, CDK1, CDC20, CCNB1, CENPF, and MAD2L1, that were related to the early diagnosis, tumour stage, and poor outcomes of HBV-HCC.

CONCLUSIONS

Five abnormally expressed hub genes of HBV-HCC are informative for early diagnosis, tumour stage determination, and poor outcome prediction.

oncoPredict: an R package for predicting in vivo or cancer patient drug response and biomarkers from cell line screening data

Cell line drug screening datasets can be utilized for a range of different drug discovery applications from drug biomarker discovery to building translational models of drug response. Previously, we described three separate methodologies to (1) correct for general levels of drug sensitivity to enable drug-specific biomarker discovery, (2) predict clinical drug response in patients and (3) associate these predictions with clinical features to perform in vivo drug biomarker discovery. Here, we unite and update these methodologies into one R package (oncoPredict) to facilitate the development and adoption of these tools. This new OncoPredict R package can be applied to various in vitro and in vivo contexts for drug and biomarker discovery.

Oncogene or tumor suppressor? Long noncoding RNAs role in patient's prognosis varies depending on disease type

Functional studies of long noncoding RNAs (lncRNAs) are often performed in the context of only a single cancer type. However, the tissue-specific expression patterns of lncRNAs raise the question of whether lncRNA associations identified in one cancer type are relevant to other cancer types. Here, we examine the relationships between the expression levels of 50 cancer-related lncRNAs and survival data from 24 types of cancer in The Cancer Genome Atlas (TCGA) with the goal of identifying prognosis related lncRNAs. Our results suggest that high expression levels of certain lncRNAs are consistently associated with worse/better survival in a number of cancers, while other lncRNAs have different prognostic roles in different types of cancer. Our analysis also identifies 20 novel unadjusted associations that have not been reported before. In addition, in low-grade glioma (LGG), prognostic-related lncRNAs are identified after conditioning on known clinical biomarker and common therapy, revealing that 2 lncRNAs, FOXP4-AS1, and NEAT1, are associated with temozolomide response-a standard-of-care in LGG. Pathway analysis suggests NF-kB/STAT3 signaling pathway enrichment in LGG patients with high NEAT1 expression and DNA repair/myc gene set enrichment in LGG patients with high expression of FOXP4-AS1. Our work demonstrates the context dependency of lncRNAs across cancer types and highlights a number of lncRNAs as potential novel cancer prognosis markers.

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.

Computationally predicting clinical drug combination efficacy with cancer cell line screens and independent drug action

Evidence has recently emerged that many clinical cancer drug combinations may derive their efficacy from independent drug action (IDA), where patients only receive benefit from the single most effective drug in a drug combination. Here we present IDACombo, an IDA based method to predict the efficacy of drug combinations using monotherapy data from high-throughput cancer cell line screens. We show that IDACombo predictions closely agree with measured drug combination efficacies both in vitro (Pearson's correlation = 0.93 when comparing predicted efficacies to measured efficacies for >5000 combinations) and in a systematically selected set of clinical trials (accuracy > 84% for predicting statistically significant improvements in patient outcomes for 26 first line therapy trials). Finally, we demonstrate how IDACombo can be used to systematically prioritize combinations for development in specific cancer settings, providing a framework for quickly translating existing monotherapy cell line data into clinically meaningful predictions of drug combination efficacy.

Evaluation of publicly available in vitro drug sensitivity models for ovarian and uterine cancer

OBJECTIVE

Publicly available data on drug sensitivity for cancer cell lines have been curated into a single, integrated database, PharmacoDB. The contributing datasets report modeled estimates of drug effect from high throughput assays. These databases have been informative for developing new broad insights, but the reliability of these data specifically for drugs used to treat ovarian and uterine cancers in related cell lines has not been reported.

METHODS

In vitro viability assays were performed on A2780, OVCAR-3, TOV-21G, and RL95-2 cells with nine drugs to produce high resolution exposure-response curves. Lab generated data were compared to publicly available datasets by IC₂₀, IC₅₀, and IC₈₀ values, and the area between the logarithmic logistic regression curves.

RESULTS

For exposure-response curve comparisons with clinically indicated drugs between lab generated and publicly available data, the majority had area-between-curves less than 20%, indicating similarity. However, 15 out of 40 of these dataset curves were incomplete as indicated by the lack of, or extrapolated, IC₅₀ value. The common ovarian and uterine cancer drug, carboplatin, exemplified this incomplete status as all of the available dataset curves were incomplete and therefore non-informative.

CONCLUSIONS

For gynecologic malignancy cell line models, experimental drug sensitivity data is comparable to the available data in PharmacoDB when exposure-response curves are complete. Incomplete exposure-response curves due to incomplete concentration ranges tested and related extrapolation of IC values can mislead individual drug/cell line pair data for downstream applications.

Machine learning approaches to drug response prediction: challenges and recent progress

Cancer is a leading cause of death worldwide. Identifying the best treatment using computational models to personalize drug response prediction holds great promise to improve patient's chances of successful recovery. Unfortunately, the computational task of predicting drug response is very challenging, partially due to the limitations of the available data and partially due to algorithmic shortcomings. The recent advances in deep learning may open a new chapter in the search for computational drug response prediction models and ultimately result in more accurate tools for therapy response. This review provides an overview of the computational challenges and advances in drug response prediction, and focuses on comparing the machine learning techniques to be of utmost practical use for clinicians and machine learning non-experts. The incorporation of new data modalities such as single-cell profiling, along with techniques that rapidly find effective drug combinations will likely be instrumental in improving cancer care.

Pharmacogenetic and pharmacogenomic discovery strategies

Genetic/genomic profiling at a single-patient level is expected to provide critical information for determining inter-individual drug toxicity and potential efficacy in cancer therapy. A better definition of cancer subtypes at a molecular level, may correspondingly complement such pharmacogenetic and pharmacogenomic approaches, for more effective personalized treatments. Current pharmacogenetic/pharmacogenomic strategies are largely based on the identification of known polymorphisms, thus limiting the discovery of novel or rarer genetic variants. Recent improvements in cost and throughput of next generation sequencing (NGS) are now making whole-genome profiling a plausible alternative for clinical procedures. Beyond classical pharmacogenetic/pharmacogenomic traits for drug metabolism, NGS screening programs of cancer genomes may lead to the identification of novel cancer-driving mutations. These may not only constitute novel therapeutic targets, but also effector determinants for metabolic pathways linked to drug metabolism. An additional advantage is that cancer NGS profiling is now leading to discovering targetable mutations, e.g., in glioblastomas and pancreatic cancers, which were originally discovered in other tumor types, thus allowing for effective repurposing of active drugs already on the market.

A simplified transposon mutagenesis method to perform phenotypic forward genetic screens in cultured cells

BACKGROUND

The introduction of genome-wide shRNA and CRISPR libraries has facilitated cell-based screens to identify loss-of-function mutations associated with a phenotype of interest. Approaches to perform analogous gain-of-function screens are less common, although some reports have utilized arrayed viral expression libraries or the CRISPR activation system. However, a variety of technical and logistical challenges make these approaches difficult for many labs to execute. In addition, genome-wide shRNA or CRISPR libraries typically contain of hundreds of thousands of individual engineered elements, and the associated complexity creates issues with replication and reproducibility for these methods.

RESULTS

Here we describe a simple, reproducible approach using the SB transposon system to perform phenotypic cell-based genetic screens. This approach employs only three plasmids to perform unbiased, whole-genome transposon mutagenesis. We also describe a ligation-mediated PCR method that can be used in conjunction with the included software tools to map raw sequence data, identify candidate genes associated with phenotypes of interest, and predict the impact of recurrent transposon insertions on candidate gene function. Finally, we demonstrate the high reproducibility of our approach by having three individuals perform independent replicates of a mutagenesis screen to identify drivers of vemurafenib resistance in cultured melanoma cells.

CONCLUSIONS

Collectively, our work establishes a facile, adaptable method that can be performed by labs of any size to perform robust, genome-wide screens to identify genes that influence phenotypes of interest.