Gene isoforms as expression-based biomarkers predictive of drug response in vitro

Likelihood-based inference under nonconvex boundary constraints

Likelihood-based inference under nonconvex constraints on model parameters has become increasingly common in biomedical research. In this paper, we establish large-sample properties of the maximum likelihood estimator when the true parameter value lies at the boundary of a nonconvex parameter space. We further derive the asymptotic distribution of the likelihood ratio test statistic under nonconvex constraints on model parameters. A general Monte Carlo procedure for generating the limiting distribution is provided. The theoretical results are demonstrated by five examples in Anderson's stereotype logistic regression model, genetic association studies, gene-environment interaction tests, cost-constrained linear regression and fairness-constrained linear regression.

Exploring the role of ITGB6: fibrosis, cancer, and other diseases

Integrin β6 (ITGB6), a member of the integrin family of proteins, is only present in epithelial tissues and frequently associates with integrin subunit αv to form transmembrane heterodimers named integrin αvβ6. Importantly, ITGB6 determines αvβ6 expression and availability. In addition to being engaged in organ fibrosis, ITGB6 is also directly linked to the emergence of cancer, periodontitis, and several potential genetic diseases. Therefore, it is of great significance to study the molecular-biological mechanism of ITGB6, which could provide novel insights for future clinical diagnosis and therapy. This review introduces the structure, distribution, and biological function of ITGB6. This review also expounds on ITGB6-related diseases, detailing the known biological effects of ITGB6.

The necroptosis signature and molecular mechanism of lung squamous cell carcinoma

BACKGROUND

Given the poor prognosis of lung squamous cell carcinoma (LUSC), the aim of this study was to screen for new prognostic biomarkers.

METHODS

The TGCA_LUSC dataset was used as the training set, and GSE73403 was used as the validation set. The genes involved in necroptosis-related pathways were acquired from the KEGG database, and the differential genes between the LUSC and normal samples were identified using the GSEA. A necroptosis signature was constructed by survival analysis, and its correlation with patient prognosis and clinical features was evaluated. The molecular characteristics and drug response associated with the necroptosis signature were also identified. The drug candidates were then validated at the cellular level.

RESULTS

The TCGA_LUSC dataset included 51 normal samples and 502 LUSC samples. The GSE73403 dataset included 69 samples. 159 genes involved in necroptosis pathways were acquired from the KEGG database, of which most showed significant differences between two groups in terms of genomic, transcriptional and methylation alterations. In particular, CHMP4C, IL1B, JAK1, PYGB and TNFRSF10B were significantly associated with the survival (p < 0.05) and were used to construct the necroptosis signature, which showed significant correlation with patient prognosis and clinical features in univariate and multivariate analyses (p < 0.05). Furthermore, CHMP4C, IL1B, JAK1 and PYGB were identified as potential targets of trametinib, selumetinib, SCH772984, PD 325901 and dasatinib. Finally, knockdown of these genes in LUSC cells increased chemosensitivity to those drugs.

CONCLUSION

We identified a necroptosis signature in LUSC that can predict prognosis and identify patients who can benefit from targeted therapies.

ACTOR: a latent Dirichlet model to compare expressed isoform proportions to a reference panel

The relative proportion of RNA isoforms expressed for a given gene has been associated with disease states in cancer, retinal diseases, and neurological disorders. Examination of relative isoform proportions can help determine biological mechanisms, but such analyses often require a per-gene investigation of splicing patterns. Leveraging large public data sets produced by genomic consortia as a reference, one can compare splicing patterns in a data set of interest with those of a reference panel in which samples are divided into distinct groups, such as tissue of origin, or disease status. We propose A latent Dirichlet model to Compare expressed isoform proportions TO a Reference panel (ACTOR), a latent Dirichlet model with Dirichlet Multinomial observations to compare expressed isoform proportions in a data set to an independent reference panel. We use a variational Bayes procedure to estimate posterior distributions for the group membership of one or more samples. Using the Genotype-Tissue Expression project as a reference data set, we evaluate ACTOR on simulated and real RNA-seq data sets to determine tissue-type classifications of genes. ACTOR is publicly available as an R package at https://github.com/mccabes292/actor.

The impact of violating the independence assumption in meta-analysis on biomarker discovery

With rapid advancements in high-throughput sequencing technologies, massive amounts of "-omics" data are now available in almost every biomedical field. Due to variance in biological models and analytic methods, findings from clinical and biological studies are often not generalizable when tested in independent cohorts. Meta-analysis, a set of statistical tools to integrate independent studies addressing similar research questions, has been proposed to improve the accuracy and robustness of new biological insights. However, it is common practice among biomarker discovery studies using preclinical pharmacogenomic data to borrow molecular profiles of cancer cell lines from one study to another, creating dependence across studies. The impact of violating the independence assumption in meta-analyses is largely unknown. In this study, we review and compare different meta-analyses to estimate variations across studies along with biomarker discoveries using preclinical pharmacogenomics data. We further evaluate the performance of conventional meta-analysis where the dependence of the effects was ignored via simulation studies. Results show that, as the number of non-independent effects increased, relative mean squared error and lower coverage probability increased. Additionally, we also assess potential bias in the estimation of effects for established meta-analysis approaches when data are duplicated and the assumption of independence is violated. Using pharmacogenomics biomarker discovery, we find that treating dependent studies as independent can substantially increase the bias of meta-analyses. Importantly, we show that violating the independence assumption decreases the generalizability of the biomarker discovery process and increases false positive results, a key challenge in precision oncology.

Expression dynamics of periodic transcripts during cancer cell cycle progression and their correlation with anticancer drug sensitivity

BACKGROUND

The cell cycle is at the center of cellular activities and is orchestrated by complex regulatory mechanisms, among which transcriptional regulation is one of the most important components. Alternative splicing dramatically expands the regulatory network by producing transcript isoforms of genes to exquisitely control the cell cycle. However, the patterns of transcript isoform expression in the cell cycle are unclear. Therapies targeting cell cycle checkpoints are commonly used as anticancer therapies, but none of them have been designed or evaluated at the alternative splicing transcript level. The utility of these transcripts as markers of cell cycle-related drug sensitivity is still unknown, and studies on the expression patterns of cell cycle-targeting drug-related transcripts are also rare.

METHODS

To explore alternative splicing patterns during cell cycle progression, we performed sequential transcriptomic assays following cell cycle synchronization in colon cancer HCT116 and breast cancer MDA-MB-231 cell lines, using flow cytometry and reference cell cycle transcripts to confirm the cell cycle phases of samples, and we developed a new algorithm to describe the periodic patterns of transcripts fluctuating during the cell cycle. Genomics of Drug Sensitivity in Cancer (GDSC) drug sensitivity datasets and Cancer Cell Line Encyclopedia (CCLE) transcript datasets were used to assess the correlation of genes and their transcript isoforms with drug sensitivity. We identified transcripts associated with typical drugs targeting cell cycle by determining correlation coefficients. Cytotoxicity assays were used to confirm the effect of ENST00000257904 against cyclin dependent kinase 4/6 (CDK4/6) inhibitors. Finally, alternative splicing transcripts associated with mitotic (M) phase arrest were analyzed using an RNA synthesis inhibition assay and transcriptome analysis.

RESULTS

We established high-resolution transcriptome datasets of synchronized cell cycle samples from colon cancer HCT116 and breast cancer MDA-MB-231 cells. The results of the cell cycle assessment showed that 43,326, 41,578 and 29,244 transcripts were found to be periodically expressed in HeLa, HCT116 and MDA-MB-231 cells, respectively, among which 1280 transcripts showed this expression pattern in all three cancer cell lines. Drug sensitivity assessments showed that a large number of these transcripts displayed a higher correlation with drug sensitivity than their corresponding genes. Cell cycle-related drug screening showed that the level of the CDK4 transcript ENST00000547281 was more significantly associated with the resistance of cells to CDK4/6 inhibitors than the level of the CDK4 reference transcript ENST00000257904. The transcriptional inhibition assay following M phase arrest further confirmed the M-phase-specific expression of the splicing transcripts. Combined with the cell cycle-related drug screening, the results also showed that a set of periodic transcripts, for example, ENST00000314392 (a dolichyl-phosphate mannosyltransferase polypeptide 2 isoform transcript), was more associated with drug sensitivity than the levels of their corresponding gene transcripts.

CONCLUSIONS

In summary, we identified a panel of cell cycle-related periodic transcripts and found that the levels of transcripts of drug target genes showed different values for predicting drug sensitivity, providing novel insights into alternative splicing-related drug development and evaluation.

Tissue-specific identification of multi-omics features for pan-cancer drug response prediction

Current statistical models for drug response prediction and biomarker identification fall short in leveraging the shared and unique information from various cancer tissues and multi-omics profiles. We developed mix-lasso model that introduces an additional sample group penalty term to capture tissue-specific effects of features on pan-cancer response prediction. The mix-lasso model takes into account both the similarity between drug responses (i.e., multi-task learning), and the heterogeneity between multi-omics data (multi-modal learning). When applied to large-scale pharmacogenomics dataset from Cancer Therapeutics Response Portal, mix-lasso enabled accurate drug response predictions and identification of tissue-specific predictive features in the presence of various degrees of missing data, drug-drug correlations, and high-dimensional and correlated genomic and molecular features that often hinder the use of statistical approaches in drug response modeling. Compared to tree lasso model, mix-lasso identified a smaller number of tissue-specific features, hence making the model more interpretable and stable for drug discovery applications.

•

Pan-cancer cell lines provide a test bench for exploring gene-drug relationships

•

Multi-omics data were integrated with pharmacological profiles for joint modeling

•

Mix-lasso identifies tissue-specific biomarkers predictive of multi-drug responses

•

Mix-lasso provides small number of stable features for drug discovery applications

Bimodal gene expression in cancer patients provides interpretable biomarkers for drug sensitivity

Identifying biomarkers predictive of cancer cell response to drug treatment constitutes one of the main challenges in precision oncology. Recent large-scale cancer pharmacogenomic studies have opened new avenues of research to develop predictive biomarkers by profiling thousands of human cancer cell lines at the molecular level and screening them with hundreds of approved drugs and experimental chemical compounds. Many studies have leveraged these data to build predictive models of response using various statistical and machine learning methods. However, a common pitfall to these methods is the lack of interpretability as to how they make predictions, hindering the clinical translation of these models. To alleviate this issue, we used the recent logic modeling approach to develop a new machine learning pipeline that explores the space of bimodally expressed genes in multiple large in vitro pharmacogenomic studies and builds multivariate, nonlinear, yet interpretable logic-based models predictive of drug response. The performance of this approach was showcased in a compendium of the three largest in vitro pharmacogenomic data sets to build robust and interpretable models for 101 drugs that span 17 drug classes with high validation rates in independent datasets. These results along with in vivo and clinical validation, support a better translation of gene expression biomarkers between model systems using bimodal gene expression.

MultiTrans: An Algorithm for Path Extraction Through Mixed Integer Linear Programming for Transcriptome Assembly

Recent advances in RNA-seq technology have made identification of expressed genes affordable, and thus boosting repaid development of transcriptomic studies. Transcriptome assembly, reconstructing all expressed transcripts from RNA-seq reads, is an essential step to understand genes, proteins, and cell functions. Transcriptome assembly remains a challenging problem due to complications in splicing variants, expression levels, uneven coverage and sequencing errors. Here, we formulate the transcriptome assembly problem as path extraction on splicing graphs (or assembly graphs), and propose a novel algorithm MultiTrans for path extraction using mixed integer linear programming. MultiTrans is able to take into consideration coverage constraints on vertices and edges, the number of paths and the paired-end information simultaneously. We benchmarked MultiTrans against two state-of-the-art transcriptome assemblers, TransLiG and rnaSPAdes. Experimental results show that MultiTrans generates more accurate transcripts compared to TransLiG (using the same splicing graphs) and rnaSPAdes (using the same assembly graphs). MultiTrans is freely available at https://github.com/jzbio/MultiTrans.

Transcript-targeted analysis reveals isoform alterations and double-hop fusions in breast cancer

Although transcriptome alteration is an essential driver of carcinogenesis, the effects of chromosomal structural alterations on the cancer transcriptome are not yet fully understood. Short-read transcript sequencing has prevented researchers from directly exploring full-length transcripts, forcing them to focus on individual splice sites. Here, we develop a pipeline for Multi-Sample long-read Transcriptome Assembly (MuSTA), which enables construction of a transcriptome from long-read sequence data. Using the constructed transcriptome as a reference, we analyze RNA extracted from 22 clinical breast cancer specimens. We identify a comprehensive set of subtype-specific and differentially used isoforms, which extended our knowledge of isoform regulation to unannotated isoforms including a short form TNS3. We also find that the exon-intron structure of fusion transcripts depends on their genomic context, and we identify double-hop fusion transcripts that are transcribed from complex structural rearrangements. For example, a double-hop fusion results in aberrant expression of an endogenous retroviral gene, ERVFRD-1, which is normally expressed exclusively in placenta and is thought to protect fetus from maternal rejection; expression is elevated in several TCGA samples with ERVFRD-1 fusions. Our analyses provide direct evidence that full-length transcript sequencing of clinical samples can add to our understanding of cancer biology and genomics in general.

Network-guided identification of cancer-selective combinatorial therapies in ovarian cancer

Each patient's cancer consists of multiple cell subpopulations that are inherently heterogeneous and may develop differing phenotypes such as drug sensitivity or resistance. A personalized treatment regimen should therefore target multiple oncoproteins in the cancer cell populations that are driving the treatment resistance or disease progression in a given patient to provide maximal therapeutic effect, while avoiding severe co-inhibition of non-malignant cells that would lead to toxic side effects. To address the intra- and inter-tumoral heterogeneity when designing combinatorial treatment regimens for cancer patients, we have implemented a machine learning-based platform to guide identification of safe and effective combinatorial treatments that selectively inhibit cancer-related dysfunctions or resistance mechanisms in individual patients. In this case study, we show how the platform enables prediction of cancer-selective drug combinations for patients with high-grade serous ovarian cancer using single-cell imaging cytometry drug response assay, combined with genome-wide transcriptomic and genetic profiles. The platform makes use of drug-target interaction networks to prioritize those combinations that warrant further preclinical testing in scarce patient-derived primary cells. During the case study in ovarian cancer patients, we investigated (i) the relative performance of various ensemble learning algorithms for drug response prediction, (ii) the use of matched single-cell RNA-sequencing data to deconvolute cell population-specific transcriptome profiles from bulk RNA-seq data, (iii) and whether multi-patient or patient-specific predictive models lead to better predictive accuracy. The general platform and the comparison results are expected to become useful for future studies that use similar predictive approaches also in other cancer types.

CircCD44 plays oncogenic roles in triple-negative breast cancer by modulating the miR-502-5p/KRAS and IGF2BP2/Myc axes

BACKGROUND

Emerging studies have revealed the potent functions of circRNAs in breast cancer tumorigenesis. However, the biogenesis, biofunction and mechanism of circRNAs in triple-negative breast cancer (TNBC) are largely unknown.

METHODS

High-throughput RNA sequencing was applied to identify dysregulated circRNAs in TNBCs and paired normal tissues. RNA pulldown and luciferase assays were performed to investigate the interaction between circular CD44 (circCD44, also annotated as hsa_circ_0021735) and miR-502-5p. RNA pulldown and RIP assays were used to investigate the interaction between circCD44 and IGF2BP2. Cell viability, colony formation, migration/invasion assays and in vivo tumorigenesis were used to investigate circCD44 biological functions.

RESULTS

CircCD44 is an uncharacterized circRNA, which is highly expressed in TNBC, and its expression is negatively correlated with the prognosis of TNBC patients. CircCD44 promotes TNBC proliferation, migration, invasion and tumorigenesis at least partially by sponging miR-502-5p and interacting with IGF2BP2.

CONCLUSION

Our data suggested that overexpressed circCD44 promotes TNBC progression. CircCD44 is potentially a novel diagnostic and therapeutic marker for TNBC patients.

Gα15 in early onset of pancreatic ductal adenocarcinoma

The GNA15 gene is ectopically expressed in human pancreatic ductal adenocarcinoma cancer cells. The encoded Gα15 protein can promiscuously redirect GPCR signaling toward pathways with oncogenic potential. We sought to describe the distribution of GNA15 in adenocarcinoma from human pancreatic specimens and to analyze the mechanism driving abnormal expression and the consequences on signaling and clinical follow-up. We detected GNA15 expression in pre-neoplastic pancreatic lesions and throughout progression. The analysis of biological data sets, primary and xenografted human tumor samples, and clinical follow-up shows that elevated expression is associated with poor prognosis for GNA15, but not any other GNA gene. Demethylation of the 5′ GNA15 promoter region was associated with ectopic expression of Gα15 in pancreatic neoplastic cells, but not in adjacent dysplastic or non-transformed tissue. Down-modulation of Gα15 by shRNA or CRISPR/Cas9 affected oncogenic signaling, and reduced adenocarcimoma cell motility and invasiveness. We conclude that de novo expression of wild-type GNA15 characterizes transformed pancreatic cells. The methylation pattern of GNA15 changes in preneoplastic lesions coincident with the release a transcriptional blockade that allows ectopic expression to persist throughout PDAC progression. Elevated GNA15 mRNA correlates with poor prognosis. In addition, ectopic Gα15 signaling provides an unprecedented mechanism in the early steps of pancreas carcinogenesis distinct from classical G protein oncogenic mutations described previously in GNAS and GNAQ/GNA11.

Alternative RNA Splicing-The Trojan Horse of Cancer Cells in Chemotherapy

Almost all transcribed human genes undergo alternative RNA splicing, which increases the diversity of the coding and non-coding cellular landscape. The resultant gene products might have distinctly different and, in some cases, even opposite functions. Therefore, the abnormal regulation of alternative splicing plays a crucial role in malignant transformation, development, and progression, a fact supported by the distinct splicing profiles identified in both healthy and tumor cells. Drug resistance, resulting in treatment failure, still remains a major challenge for current cancer therapy. Furthermore, tumor cells often take advantage of aberrant RNA splicing to overcome the toxicity of the administered chemotherapeutic agents. Thus, deciphering the alternative RNA splicing variants in tumor cells would provide opportunities for designing novel therapeutics combating cancer more efficiently. In the present review, we provide a comprehensive outline of the recent findings in alternative splicing in the most common neoplasms, including lung, breast, prostate, head and neck, glioma, colon, and blood malignancies. Molecular mechanisms developed by cancer cells to promote oncogenesis as well as to evade anticancer drug treatment and the subsequent chemotherapy failure are also discussed. Taken together, these findings offer novel opportunities for future studies and the development of targeted therapy for cancer-specific splicing variants.

Machine learning in the prediction of cancer therapy

Resistance to therapy remains a major cause of cancer treatment failures, resulting in many cancer-related deaths. Resistance can occur at any time during the treatment, even at the beginning. The current treatment plan is dependent mainly on cancer subtypes and the presence of genetic mutations. Evidently, the presence of a genetic mutation does not always predict the therapeutic response and can vary for different cancer subtypes. Therefore, there is an unmet need for predictive models to match a cancer patient with a specific drug or drug combination. Recent advancements in predictive models using artificial intelligence have shown great promise in preclinical settings. However, despite massive improvements in computational power, building clinically useable models remains challenging due to a lack of clinically meaningful pharmacogenomic data. In this review, we provide an overview of recent advancements in therapeutic response prediction using machine learning, which is the most widely used branch of artificial intelligence. We describe the basics of machine learning algorithms, illustrate their use, and highlight the current challenges in therapy response prediction for clinical practice.

Differential Splicing of Skipped-exons Predicts Drug Response in Cancer Cell Lines

Alternative splicing of pre-mRNA transcripts is an important regulatory mechanism that increases the diversity of gene products in eukaryotes. Various studies have linked specific transcript isoforms to altered drug response in cancer; however, few algorithms have incorporated splicing information into drug response prediction. In this study, we evaluated whether basal-level splicing information could be used to predict drug sensitivity by constructing doxorubicin-sensitivity classification models with splicing and expression data. We detailed splicing differences between sensitive and resistant cell lines by implementing quasi-binomial generalized linear modeling (QBGLM) and found altered inclusion of 277 skipped exons. We additionally conducted RNA-binding protein (RBP) binding motif enrichment and differential expression analysis to characterize cis- and trans-acting elements that potentially influence doxorubicin response-mediating splicing alterations. Our results showed that a classification model built with skipped exon data exhibited strong predictive power. We discovered an association between differentially spliced events and epithelial-mesenchymal transition (EMT) and observed motif enrichment, as well as differential expression of RBFOX and ELAVL RBP family members. Our work demonstrates the potential of incorporating splicing data into drug response algorithms and the utility of a QBGLM approach for fast, scalable identification of relevant splicing differences between large groups of samples.

Alternative Splicing: Expanding the Landscape of Cancer Biomarkers and Therapeutics

Alternative splicing (AS) is a critical post-transcriptional regulatory mechanism used by more than 95% of transcribed human genes and responsible for structural transcript variation and proteome diversity. In the past decade, genome-wide transcriptome sequencing has revealed that AS is tightly regulated in a tissue- and developmental stage-specific manner, and also frequently dysregulated in multiple human cancer types. It is currently recognized that splicing defects, including genetic alterations in the spliced gene, altered expression of both core components or regulators of the precursor messenger RNA (pre-mRNA) splicing machinery, or both, are major drivers of tumorigenesis. Hence, in this review we provide an overview of our current understanding of splicing alterations in cancer, and emphasize the need to further explore the cancer-specific splicing programs in order to obtain new insights in oncology. Furthermore, we also discuss the recent advances in the identification of dysregulated splicing signatures on a genome-wide scale and their potential use as biomarkers. Finally, we highlight the therapeutic opportunities arising from dysregulated splicing and summarize the current approaches to therapeutically target AS in cancer.

Feature set optimization in biomarker discovery from genome-scale data

MOTIVATION

Omics technologies have the potential to facilitate the discovery of new biomarkers. However, only few omics-derived biomarkers have been successfully translated into clinical applications to date. Feature selection is a crucial step in this process that identifies small sets of features with high predictive power. Models consisting of a limited number of features are not only more robust in analytical terms, but also ensure cost effectiveness and clinical translatability of new biomarker panels. Here we introduce GARBO, a novel multi-island adaptive genetic algorithm to simultaneously optimize accuracy and set size in omics-driven biomarker discovery problems.

RESULTS

Compared to existing methods, GARBO enables the identification of biomarker sets that best optimize the trade-off between classification accuracy and number of biomarkers. We tested GARBO and six alternative selection methods with two high relevant topics in precision medicine: cancer patient stratification and drug sensitivity prediction. We found multivariate biomarker models from different omics data types such as mRNA, miRNA, copy number variation, mutation and DNA methylation. The top performing models were evaluated by using two different strategies: the Pareto-based selection, and the weighted sum between accuracy and set size (w = 0.5). Pareto-based preferences show the ability of the proposed algorithm to search minimal subsets of relevant features that can be used to model accurate random forest-based classification systems. Moreover, GARBO systematically identified, on larger omics data types, such as gene expression and DNA methylation, biomarker panels exhibiting higher classification accuracy or employing a number of features much lower than those discovered with other methods. These results were confirmed on independent datasets.

AVAILABILITY AND IMPLEMENTATION

github.com/Greco-Lab/GARBO.

CONTACT

dario.greco@tuni.fi.

SUPPLEMENTARY INFORMATION

Supplementary data are available at Bioinformatics online.

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.

Dr.VAE: improving drug response prediction via modeling of drug perturbation effects

Motivation

Individualized drug response prediction is a fundamental part of personalized medicine for cancer. Great effort has been made to discover biomarkers or to develop machine learning methods for accurate drug response prediction in cancers. Incorporating prior knowledge of biological systems into these methods is a promising avenue to improve prediction performance. High-throughput cell line assays of drug-induced transcriptomic perturbation effects are a prior knowledge that has not been fully incorporated into a drug response prediction model yet.

Results

We introduce a unified probabilistic approach, Drug Response Variational Autoencoder (Dr.VAE), that simultaneously models both drug response in terms of viability and transcriptomic perturbations. Dr.VAE is a deep generative model based on variational autoencoders. Our experimental results showed Dr.VAE to do as well or outperform standard classification methods for 23 out of 26 tested Food and Drug Administration-approved drugs. In a series of ablation experiments we showed that the observed improvement of Dr.VAE can be credited to the incorporation of drug-induced perturbation effects with joint modeling of treatment sensitivity.

Availability and implementation

Processed data and software implementation using PyTorch (Paszke et al., 2017) are available at: https://github.com/rampasek/DrVAE.

Supplementary information

Supplementary data are available at Bioinformatics online.

Integrated individually electrochemical array for simultaneously detecting multiple Alzheimer's biomarkers

Simultaneous detection of multiple biomarkers is benefit for reducing the detection cycles, avoiding the false-positive signals, and providing the cross validation, which provide the opportunity to understand the pathogenic mechanisms and achieve precise early diagnosis. Here, we demonstrate the mini-pillar-based individual electrochemical array for simultaneous detection of multiple biomarkers. On such platform, the mini-pillar could confine the microdroplet as individual and open-channel microreactor, which is extremely helpful for reducing reagent consumption and extracting internal information, and the electrodes array embedded in mini-pillar are integrated on one side to achieve multiple and simultaneous electrochemical sensing. The introduction of gold nanodendrites by electrodeposition has greatly enhanced sensitivity via improving probe-binding capacity and response signals. Sensitive and selective detection of multiple Alzheimer's biomarkers including Tau, ApoE4, Amyloid-β and miRNA-101 on such mini-pillar-based biosensor is also achieved. Such biosensor platform with the advantages of high-yield, high sensitivity, low-waste and multiple signals output shows great promise in sensing multiple biomolecules for disease diagnosis and health monitoring.

Platform-integrated mRNA isoform quantification

MOTIVATION

Accurate estimation of transcript isoform abundance is critical for downstream transcriptome analyses and can lead to precise molecular mechanisms for understanding complex human diseases, like cancer. Simplex mRNA Sequencing (RNA-Seq) based isoform quantification approaches are facing the challenges of inherent sampling bias and unidentifiable read origins. A large-scale experiment shows that the consistency between RNA-Seq and other mRNA quantification platforms is relatively low at the isoform level compared to the gene level. In this project, we developed a platform-integrated model for transcript quantification (IntMTQ) to improve the performance of RNA-Seq on isoform expression estimation. IntMTQ, which benefits from the mRNA expressions reported by the other platforms, provides more precise RNA-Seq-based isoform quantification and leads to more accurate molecular signatures for disease phenotype prediction.

RESULTS

In the experiments to assess the quality of isoform expression estimated by IntMTQ, we designed three tasks for clustering and classification of 46 cancer cell lines with four different mRNA quantification platforms, including newly developed NanoString's nCounter technology. The results demonstrate that the isoform expressions learned by IntMTQ consistently provide more and better molecular features for downstream analyses compared with five baseline algorithms which consider RNA-Seq data only. An independent RT-qPCR experiment on seven genes in twelve cancer cell lines showed that the IntMTQ improved overall transcript quantification. The platform-integrated algorithms could be applied to large-scale cancer studies, such as The Cancer Genome Atlas (TCGA), with both RNA-Seq and array-based platforms available.

AVAILABILITY AND IMPLEMENTATION

Source code is available at: https://github.com/CompbioLabUcf/IntMTQ.

SUPPLEMENTARY INFORMATION

Supplementary data are available at Bioinformatics online.

Discrete functional and mechanistic roles of chromodomain Y-like 2 (CDYL2) transcript variants in breast cancer growth and metastasis

Rationale: Chromodomain Y-like 2 (CDYL2) is a member of the CDY gene family involved in spermatogenesis, but its role in human cancer has not been reported. Analyses of publicly available databases demonstrate that CDYL2 is abundantly expressed in breast tumors. However, whether CDYL2 is involved in breast cancer progression remains unknown.

Methods: Quantitative real-time PCR and immunoblotting assays were used to determine the expression levels of CDYL2 transcript variants in breast cancer cell lines and primary breast tumors. The effect of CDYL2 transcript variants on the malignant phenotypes of breast cancer cells was examined through in vitro and in vivo assays. Immunofluorescent staining, RNA-seq, ATAC-seq, and ChIP-qPCR were used to investigate the underlying mechanisms behind the aforementioned observations.

Results: Here we show that CDYL2 generated four transcript variants, named CDYL2a-CDYL2d. CDYL2a and CDYL2b were the predominant variants expressed in breast cancer cell lines and breast tumors and exerted strikingly discrete functions in breast cancer growth and metastasis. CDYL2a was upregulated in the majority of the breast cancer cell lines and tumors, and promoted breast cancer cell proliferation, colony formation in vitro, and tumorigenesis in xenografts. In contrast, CDYL2b was mainly expressed in luminal- and HER2-positive types of breast cancer cell lines and tumors, and suppressed the migratory, invasive, and metastatic potential of breast cancer cells in vitro and in vivo. Mechanistically, CDYL2a partially localized to SC35-positive nuclear speckles and promoted alternative splicing of a subset of target genes, including FIP1L1, NKTR, and ADD3 by exon skipping. Elimination of full-length FIP1L1, NKTR, and ADD3 rescued the impaired cell proliferation through CDYL2a depletion. In contrast, CDYL2b localized to heterochromatin and transcriptionally repressed several metastasis-promoting genes, including HPSE, HLA-F, and SELL. Restoration of HPSE, HLA-F, or SELL expression in CDYL2b-overexpressing cells attenuated the ability of CDYL2b to suppress breast cancer cell migration and invasion.

Conclusions: Collectively, these findings establish an isoform-specific function of CDYL2 in breast cancer development and progression and highlight that pharmacological inhibition of the CDYL2a, but not the CDYL2b, isoform may be an effective strategy for breast cancer therapy.

In silico analysis of alternative splicing on drug-target gene interactions

Identifying and evaluating the right target are the most important factors in early drug discovery phase. Most studies focus on one protein ignoring the multiple splice-variant or protein-isoforms, which might contribute to unexpected therapeutic activity or adverse side effects. Here, we present computational analysis of cancer drug-target interactions affected by alternative splicing. By integrating information from publicly available databases, we curated 883 FDA approved or investigational stage small molecule cancer drugs that target 1,434 different genes, with an average of 5.22 protein isoforms per gene. Of these, 618 genes have ≥5 annotated protein-isoforms. By analyzing the interactions with binding pocket information, we found that 76% of drugs either miss a potential target isoform or target other isoforms with varied expression in multiple normal tissues. We present sequence and structure level alignments at isoform-level and make this information publicly available for all the curated drugs. Structure-level analysis showed ligand binding pocket architectures differences in size, shape and electrostatic parameters between isoforms. Our results emphasize how potentially important isoform-level interactions could be missed by solely focusing on the canonical isoform, and suggest that on- and off-target effects at isoform-level should be investigated to enhance the productivity of drug-discovery research.

Network-based method for drug target discovery at the isoform level

Identification of primary targets associated with phenotypes can facilitate exploration of the underlying molecular mechanisms of compounds and optimization of the structures of promising drugs. However, the literature reports limited effort to identify the target major isoform of a single known target gene. The majority of genes generate multiple transcripts that are translated into proteins that may carry out distinct and even opposing biological functions through alternative splicing. In addition, isoform expression is dynamic and varies depending on the developmental stage and cell type. To identify target major isoforms, we integrated a breast cancer type-specific isoform coexpression network with gene perturbation signatures in the MCF7 cell line in the Connectivity Map database using the ‘shortest path’ drug target prioritization method. We used a leukemia cancer network and differential expression data for drugs in the HL-60 cell line to test the robustness of the detection algorithm for target major isoforms. We further analyzed the properties of target major isoforms for each multi-isoform gene using pharmacogenomic datasets, proteomic data and the principal isoforms defined by the APPRIS and STRING datasets. Then, we tested our predictions for the most promising target major protein isoforms of DNMT1, MGEA5 and P4HB4 based on expression data and topological features in the coexpression network. Interestingly, these isoforms are not annotated as principal isoforms in APPRIS. Lastly, we tested the affinity of the target major isoform of MGEA5 for streptozocin through in silico docking. Our findings will pave the way for more effective and targeted therapies via studies of drug targets at the isoform level.

Creating reproducible pharmacogenomic analysis pipelines

The field of pharmacogenomics presents great challenges for researchers that are willing to make their studies reproducible and shareable. This is attributed to the generation of large volumes of high-throughput multimodal data, and the lack of standardized workflows that are robust, scalable, and flexible to perform large-scale analyses. To address this issue, we developed pharmacogenomic workflows in the Common Workflow Language to process two breast cancer datasets in a reproducible and transparent manner. Our pipelines combine both pharmacological and molecular profiles into a portable data object that can be used for future analyses in cancer research. Our data objects and workflows are shared on Harvard Dataverse and Code Ocean where they have been assigned a unique Digital Object Identifier, providing a level of data provenance and a persistent location to access and share our data with the community.

RNA Sequencing of the NCI-60: Integration into CellMiner and CellMiner CDB

CellMiner (http://discover.nci.nih.gov/cellminer) and CellMinerCDB (https://discover.nci.nih.gov/cellminercdb/) are web-based applications for mining publicly available genomic, molecular, and pharmacologic datasets of human cancer cell lines including the NCI-60, Cancer Cell Line Encyclopedia, Genomics of Drug Sensitivity in Cancer, Cancer Therapeutics Response Portal, NCI/DTP small cell lung cancer, and NCI Almanac cell line sets. Here, we introduce our RNA sequencing (RNA-seq) data for the NCI-60 and their access and integration with the other databases. Correlation to transcript microarray expression levels for identical genes and identical cell lines across CellMinerCDB demonstrates the high quality of these new RNA-seq data. We provide composite and isoform transcript expression data and demonstrate diversity in isoform composition for individual cancer- and pharmacologically relevant genes, including HRAS, PTEN, EGFR, RAD51, ALKBH2, BRCA1, ERBB2, TP53, FGFR2, and CTNND1. We reveal cell-specific differences in the overall levels of isoforms and show their linkage to expression of RNA processing and splicing genes as well as resultant alterations in cancer and pharmacologic gene sets. Gene-drug pairings linked by pathways or functions show specific correlations to isoforms compared with composite gene expression, including ALKBH2-benzaldehyde, AKT3-vandetanib, BCR-imatinib, CDK1 and 20-palbociclib, CASP1-imexon, and FGFR3-pazopanib. Loss of MUC1 20 amino acid variable number tandem repeats, which is used to elicit immune response, and the presence of the androgen receptor AR-V4 and -V7 isoforms in all NCI-60 tissue of origin types demonstrate translational relevance. In summary, we introduce RNA-seq data to our CellMiner and CellMinerCDB web applications, allowing their exploration for both research and translational purposes. SIGNIFICANCE: The current study provides RNA sequencing data for the NCI-60 cell lines made accessible through both CellMiner and CellMinerCDB and is an important pharmacogenomics resource for the field.

Alternative splicing, RNA-seq and drug discovery

Alternative splicing, hereafter referred to as AS, is an essential component of gene expression regulation that contributes to the diversity of proteomes. Recent developments in RNA sequencing (RNA-seq) technologies, combined with the advent of computational tools, have enabled transcriptome-wide studies of AS at an unprecedented scale and resolution. RNA mis-splicing can cause human disease, and to target alternative splicing has led to the development of novel therapeutics. Splice variants diversify the repertoire of biomarkers and functionally contribute to drug resistance. Our expanding knowledge of AS variation in human populations holds great promise for improving disease diagnoses and ultimately patient care in the era of sequencing and precision medicine.

TranscriptAchilles: a genome-wide platform to predict isoform biomarkers of gene essentiality in cancer

BACKGROUND

Aberrant alternative splicing plays a key role in cancer development. In recent years, alternative splicing has been used as a prognosis biomarker, a therapy response biomarker, and even as a therapeutic target. Next-generation RNA sequencing has an unprecedented potential to measure the transcriptome. However, due to the complexity of dealing with isoforms, the scientific community has not sufficiently exploited this valuable resource in precision medicine.

FINDINGS

We present TranscriptAchilles, the first large-scale tool to predict transcript biomarkers associated with gene essentiality in cancer. This application integrates 412 loss-of-function RNA interference screens of >17,000 genes, together with their corresponding whole-transcriptome expression profiling. Using this tool, we have studied which are the cancer subtypes for which alternative splicing plays a significant role to state gene essentiality. In addition, we include a case study of renal cell carcinoma that shows the biological soundness of the results. The databases, the source code, and a guide to build the platform within a Docker container are available at GitLab. The application is also available online.

CONCLUSIONS

TranscriptAchilles provides a user-friendly web interface to identify transcript or gene biomarkers of gene essentiality, which could be used as a starting point for a drug development project. This approach opens a wide range of translational applications in cancer.

Pan-Cancer Analyses Reveal Genomic Features of FOXM1 Overexpression in Cancer

FOXM1 is frequently overexpressed in cancer, but this has not been studied in a comprehensive manner. We utilized genotype-tissue expression (GTEx) normal and The Cancer Genome Atlas (TCGA) tumor data to define FOXM1 expression, including its isoforms, and to determine the genetic alterations that promote FOXM1 expression in cancer. Additionally, we used human fallopian tube epithelial (FTE) cells to dissect the role of Retinoblastoma (Rb)-E2F and Cyclin E1 in FOXM1 regulation, and a novel human embryonic kidney cell (HEK293T) CRISPR FOXM1 knockout model to define isoform-specific transcriptional programs. FOXM1 expression, at the mRNA and protein level, was significantly elevated in tumors with FOXM1 amplification, p53 inactivation, and Rb-E2F deregulation. FOXM1 expression was remarkably high in testicular germ cell tumors (TGCT), high-grade serous ovarian cancer (HGSC), and basal breast cancer (BBC). FOXM1 expression in cancer was associated with genomic instability, as measured using aneuploidy signatures. FTE models confirmed a role for Rb-E2F signaling in FOXM1 regulation and in particular identified Cyclin E1 as a novel inducer of FOXM1 expression. Among the three FOXM1 isoforms, FOXM1c showed the highest expression in normal and tumor tissues and cancer cell lines. The CRISPR knockout model demonstrated that FOXM1b and FOXM1c are transcriptionally active, while FOXM1a is not. Finally, we were unable to confirm the existence of a FOXM1 auto-regulatory loop. This study provides significant and novel information regarding the frequency, causes, and consequences of elevated FOXM1 expression in human cancer.

Hub genes in a pan-cancer co-expression network show potential for predicting drug responses

Background: The topological analysis of networks extracted from different types of "omics" data is a useful strategy for characterizing biologically meaningful properties of the complex systems underlying these networks. In particular, the biological significance of highly connected genes in diverse molecular networks has been previously determined using data from several model organisms and phenotypes. Despite such insights, the predictive potential of candidate hubs in gene co-expression networks in the specific context of cancer-related drug experiments remains to be deeply investigated. The examination of such associations may offer opportunities for the accurate prediction of anticancer drug responses.  Methods: Here, we address this problem by: a) analyzing a co-expression network obtained from thousands of cancer cell lines, b) detecting significant network hubs, and c) assessing their capacity to predict drug sensitivity using data from thousands of drug experiments. We investigated the prediction capability of those genes using a multiple linear regression model, independent datasets, comparisons with other models and our own in vitro experiments. Results: These analyses led to the identification of 47 hub genes, which are implicated in a diverse range of cancer-relevant processes and pathways. Overall, encouraging agreements between predicted and observed drug sensitivities were observed in public datasets, as well as in our in vitro validations for four glioblastoma cell lines and four drugs. To facilitate further research, we share our hub-based drug sensitivity prediction model as an online tool. Conclusions: Our research shows that co-expression network hubs are biologically interesting and exhibit potential for predicting drug responses in vitro. These findings motivate further investigations about the relevance and application of our unbiased discovery approach in pre-clinical, translationally-oriented research.

Hub genes in a pan-cancer co-expression network show potential for predicting drug responses

Background: The topological analysis of networks extracted from different types of "omics" data is a useful strategy for characterizing biologically meaningful properties of the complex systems underlying these networks. In particular, the biological significance of highly connected genes in diverse molecular networks has been previously determined using data from several model organisms and phenotypes. Despite such insights, the predictive potential of candidate hubs in gene co-expression networks in the specific context of cancer-related drug experiments remains to be deeply investigated. The examination of such associations may offer opportunities for the accurate prediction of anticancer drug responses.  Methods: Here, we address this problem by: a) analyzing a co-expression network obtained from thousands of cancer cell lines, b) detecting significant network hubs, and c) assessing their capacity to predict drug sensitivity using data from thousands of drug experiments. We investigated the prediction capability of those genes using a multiple linear regression model, independent datasets, comparisons with other models and our own in vitro experiments. Results: These analyses led to the identification of 47 hub genes, which are implicated in a diverse range of cancer-relevant processes and pathways. Overall, encouraging agreements between predicted and observed drug sensitivities were observed in public datasets, as well as in our in vitro validations for four glioblastoma cell lines and four drugs. To facilitate further research, we share our hub-based drug sensitivity prediction model as an online tool. Conclusions: Our research shows that co-expression network hubs are biologically interesting and exhibit potential for predicting drug responses in vitro. These findings motivate further investigations about the relevance and application of our unbiased discovery approach in pre-clinical, translationally-oriented research.

Machine learning and feature selection for drug response prediction in precision oncology applications

In-depth modeling of the complex interplay among multiple omics data measured from cancer cell lines or patient tumors is providing new opportunities toward identification of tailored therapies for individual cancer patients. Supervised machine learning algorithms are increasingly being applied to the omics profiles as they enable integrative analyses among the high-dimensional data sets, as well as personalized predictions of therapy responses using multi-omics panels of response-predictive biomarkers identified through feature selection and cross-validation. However, technical variability and frequent missingness in input "big data" require the application of dedicated data preprocessing pipelines that often lead to some loss of information and compressed view of the biological signal. We describe here the state-of-the-art machine learning methods for anti-cancer drug response modeling and prediction and give our perspective on further opportunities to make better use of high-dimensional multi-omics profiles along with knowledge about cancer pathways targeted by anti-cancer compounds when predicting their phenotypic responses.

PATRI, a Genomics Data Integration Tool for Biomarker Discovery

The availability of genomic datasets in association with clinical, phenotypic, and drug sensitivity information represents an invaluable source for potential therapeutic applications, supporting the identification of new drug sensitivity biomarkers and pharmacological targets. Drug discovery and precision oncology can largely benefit from the integration of treatment molecular discriminants obtained from cell line models and clinical tumor samples; however this task demands comprehensive analysis approaches for the discovery of underlying data connections. Here we introduce PATRI (Platform for the Analysis of TRanslational Integrated data), a standalone tool accessible through a user-friendly graphical interface, conceived for the identification of treatment sensitivity biomarkers from user-provided genomics data, associated with information on sample characteristics. PATRI streamlines a translational analysis workflow: first, baseline genomics signatures are statistically identified, differentiating treatment sensitive from resistant preclinical models; then, these signatures are used for the prediction of treatment sensitivity in clinical samples, via random forest categorization of clinical genomics datasets and statistical evaluation of the relative phenotypic features. The same workflow can also be applied across distinct clinical datasets. The ease of use of the PATRI tool is illustrated with validation analysis examples, performed with sensitivity data for drug treatments with known molecular discriminants.

Modulation of alternative splicing induced by paclitaxel in human lung cancer

Paclitaxel is utilized as the first-line chemotherapeutic regimen for the majority of advanced non-small-cell lung carcinoma. However, whether paclitaxel could suppress cancer progression through modulating RNA alternative splicing remains largely unknown. Here, we demonstrated the effects of paclitaxel on cell proliferation inhibition, cell cycle arrest, and apoptosis. Mechanistically, paclitaxel leads to transcriptional alteration of networks involved in DNA replication and repair, chromosome segregation, chromatin silencing at rDNA, and mitosis at the transcriptional level. Moreover, paclitaxel regulates a number of cancer-associated RNA alternative splicing events, including genes involved in cellular response to DNA damage stimulus, preassembly of GPI anchor in ER membrane, transcription, and DNA repair. In particular, paclitaxel modulates the splicing of ECT2, a key factor involved in the regulation of cytokinesis. Briefly, paclitaxel favors the production of ECT2-S, the short splicing isoforms of ECT2, thereby inhibiting cancer cell proliferation. Our study provides mechanistic insights of paclitaxel on RNA alternative splicing regulation, thus to offer a potential novel route for paclitaxel to inhibit cancer progression.

Disruption of the anaphase-promoting complex confers resistance to TTK inhibitors in triple-negative breast cancer

TTK protein kinase (TTK), also known as Monopolar spindle 1 (MPS1), is a key regulator of the spindle assembly checkpoint (SAC), which functions to maintain genomic integrity. TTK has emerged as a promising therapeutic target in human cancers, including triple-negative breast cancer (TNBC). Several TTK inhibitors (TTKis) are being evaluated in clinical trials, and an understanding of the mechanisms mediating TTKi sensitivity and resistance could inform the successful development of this class of agents. We evaluated the cellular effects of the potent clinical TTKi CFI-402257 in TNBC models. CFI-402257 induced apoptosis and potentiated aneuploidy in TNBC lines by accelerating progression through mitosis and inducing mitotic segregation errors. We used genome-wide CRISPR/Cas9 screens in multiple TNBC cell lines to identify mechanisms of resistance to CFI-402257. Our functional genomic screens identified members of the anaphase-promoting complex/cyclosome (APC/C) complex, which promotes mitotic progression following inactivation of the SAC. Several screen candidates were validated to confer resistance to CFI-402257 and other TTKis using CRISPR/Cas9 and siRNA methods. These findings extend the observation that impairment of the APC/C enables cells to tolerate genomic instability caused by SAC inactivation, and support the notion that a measure of APC/C function could predict the response to TTK inhibition. Indeed, an APC/C gene expression signature is significantly associated with CFI-402257 response in breast and lung adenocarcinoma cell line panels. This expression signature, along with somatic alterations in genes involved in mitotic progression, represent potential biomarkers that could be evaluated in ongoing clinical trials of CFI-402257 or other TTKis.

Revisiting inconsistency in large pharmacogenomic studies

In 2013, we published a comparative analysis of mutation and gene expression profiles and drug sensitivity measurements for 15 drugs characterized in the 471 cancer cell lines screened in the Genomics of Drug Sensitivity in Cancer (GDSC) and Cancer Cell Line Encyclopedia (CCLE). While we found good concordance in gene expression profiles, there was substantial inconsistency in the drug responses reported by the GDSC and CCLE projects. We received extensive feedback on the comparisons that we performed. This feedback, along with the release of new data, prompted us to revisit our initial analysis. We present a new analysis using these expanded data, where we address the most significant suggestions for improvements on our published analysis - that targeted therapies and broad cytotoxic drugs should have been treated differently in assessing consistency, that consistency of both molecular profiles and drug sensitivity measurements should be compared across cell lines, and that the software analysis tools provided should have been easier to run, particularly as the GDSC and CCLE released additional data. Our re-analysis supports our previous finding that gene expression data are significantly more consistent than drug sensitivity measurements. Using new statistics to assess data consistency allowed identification of two broad effect drugs and three targeted drugs with moderate to good consistency in drug sensitivity data between GDSC and CCLE. For three other targeted drugs, there were not enough sensitive cell lines to assess the consistency of the pharmacological profiles. We found evidence of inconsistencies in pharmacological phenotypes for the remaining eight drugs. Overall, our findings suggest that the drug sensitivity data in GDSC and CCLE continue to present challenges for robust biomarker discovery. This re-analysis provides additional support for the argument that experimental standardization and validation of pharmacogenomic response will be necessary to advance the broad use of large pharmacogenomic screens.

Revisiting inconsistency in large pharmacogenomic studies

In 2013, we published a comparative analysis mutation and gene expression profiles and drug sensitivity measurements for 15 drugs characterized in the 471 cancer cell lines screened in the Genomics of Drug Sensitivity in Cancer (GDSC) and Cancer Cell Line Encyclopedia (CCLE). While we found good concordance in gene expression profiles, there was substantial inconsistency in the drug responses reported by the GDSC and CCLE projects. We received extensive feedback on the comparisons that we performed. This feedback, along with the release of new data, prompted us to revisit our initial analysis. Here we present a new analysis using these expanded data in which we address the most significant suggestions for improvements on our published analysis - that targeted therapies and broad cytotoxic drugs should have been treated differently in assessing consistency, that consistency of both molecular profiles and drug sensitivity measurements should both be compared across cell lines, and that the software analysis tools we provided should have been easier to run, particularly as the GDSC and CCLE released additional data.             Our re-analysis supports our previous finding that gene expression data are significantly more consistent than drug sensitivity measurements. The use of new statistics to assess data consistency allowed us to identify two broad effect drugs and three targeted drugs with moderate to good consistency in drug sensitivity data between GDSC and CCLE. For three other targeted drugs, there were not enough sensitive cell lines to assess the consistency of the pharmacological profiles. We found evidence of inconsistencies in pharmacological phenotypes for the remaining eight drugs.             Overall, our findings suggest that the drug sensitivity data in GDSC and CCLE continue to present challenges for robust biomarker discovery. This re-analysis provides additional support for the argument that experimental standardization and validation of pharmacogenomic response will be necessary to advance the broad use of large pharmacogenomic screens.