Genomic data integration by WON-PARAFAC identifies interpretable factors for predicting drug-sensitivity in vivo

Abstract Integrative analyses that summarize and link molecular data to treatment sensitivity are crucial to capture the biological complexity which is essential to further precision medicine. We introduce Weighted Orthogonal Nonnegative parallel factor analysis (WON-PARAFAC), a data integration method that identifies sparse and interpretable factors. WON-PARAFAC summarizes the GDSC1000 cell line compendium in 130 factors. We interpret the factors based on their association with recurrent molecular alterations, pathway enrichment, cancer type, and drug-response. Crucially, the cell line derived factors capture the majority of the relevant biological variation in Patient-Derived Xenograft (PDX) models, strongly suggesting our factors capture invariant and generalizable aspects of cancer biology. Furthermore, drug response in cell lines is better and more consistently translated to PDXs using factor-based predictors as compared to raw feature-based predictors. WON-PARAFAC efficiently summarizes and integrates multiway high-dimensional genomic data and enhances translatability of drug response prediction from cell lines to patient-derived xenografts.

Download Full-text

Predicting breast cancer drug response using a multiple-layer cell line drug response network model

BMC Cancer ◽

10.1186/s12885-021-08359-6 ◽

2021 ◽

Vol 21 (1) ◽

Author(s):

Shujun Huang ◽

Pingzhao Hu ◽

Ted M. Lakowski

Keyword(s):

Cell Line ◽

Cell Lines ◽

Drug Response ◽

Expression Profiles ◽

Pearson Correlation ◽

Gene Expression Profiles ◽

Response Prediction ◽

Layer Cell ◽

Response Network ◽

Line Drug

Abstract Background Predicting patient drug response based on a patient’s molecular profile is one of the key goals of precision medicine in breast cancer (BC). Multiple drug response prediction models have been developed to address this problem. However, most of them were developed to make sensitivity predictions for multiple single drugs within cell lines from various cancer types instead of a single cancer type, do not take into account drug properties, and have not been validated in cancer patient-derived data. Among the multi-omics data, gene expression profiles have been shown to be the most informative data for drug response prediction. However, these models were often developed with individual genes. Therefore, this study aimed to develop a drug response prediction model for BC using multiple data types from both cell lines and drugs. Methods We first collected the baseline gene expression profiles of 49 BC cell lines along with IC50 values for 220 drugs tested in these cell lines from Genomics of Drug Sensitivity in Cancer (GDSC). Using these data, we developed a multiple-layer cell line-drug response network (ML-CDN2) by integrating a one-layer cell line similarity network based on the pathway activity profiles and a three-layer drug similarity network based on the drug structures, targets, and pan-cancer IC50 profiles. We further used ML-CDN2 to predict the drug response for new BC cell lines or patient-derived samples. Results ML-CDN2 demonstrated a good predictive performance, with the Pearson correlation coefficient between the observed and predicted IC50 values for all GDSC cell line-drug pairs of 0.873. Also, ML-CDN2 showed a good performance when used to predict drug response in new BC cell lines from the Cancer Cell Line Encyclopedia (CCLE), with a Pearson correlation coefficient of 0.718. Moreover, we found that the cell line-derived ML-CDN2 model could be applied to predict drug response in the BC patient-derived samples from The Cancer Genome Atlas (TCGA). Conclusions The ML-CDN2 model was built to predict BC drug response using comprehensive information from both cell lines and drugs. Compared with existing methods, it has the potential to predict the drug response for BC patient-derived samples.

Download Full-text

Application of network diffusion approaches to drug screenings: A perspective on multilayered networks derived from cell lines and drugs

10.7287/peerj.preprints.3339 ◽

2017 ◽

Author(s):

Vigneshwari Subramanian ◽

Bence Szalai ◽

Luis Tobalina ◽

Julio Saez-Rodriguez

Keyword(s):

Cell Line ◽

Cell Lines ◽

Protein Interactions ◽

Drug Response ◽

Prediction Models ◽

Response Prediction ◽

Single Type ◽

Disease Genes ◽

Cell Line Cell ◽

Network Diffusion

Network diffusion approaches are frequently used for identifying the relevant disease genes and for prioritizing the genes for drug sensitivity predictions. Majority of these studies rely on networks representing a single type of information. However, using multiplex heterogeneous networks (networks with multiple interconnected layers) is much more informative and helps to understand the global topology. We built a multi-layered network that incorporates information on protein-protein interactions, drug-drug similarities, cell line-cell line similarities and co-expressed genes. We applied Random Walk with Restart algorithm to investigate the interactions between drugs, targets and cancer cell lines. Results of ANOVA models show that these prioritized genes are among the most significant ones that relate to drug response. Moreover, the predictive power of the drug response prediction models built using the gene expression data of only the top ranked genes is similar to the models built using all the available genes. Taken together, the results confirm that the multiplex heterogeneous network-based approach is efficient in identifying the most significant genes associated with drug response.

Download Full-text

Application of network diffusion approaches to drug screenings: A perspective on multilayered networks derived from cell lines and drugs

10.7287/peerj.preprints.3339v1 ◽

2017 ◽

Author(s):

Vigneshwari Subramanian ◽

Bence Szalai ◽

Luis Tobalina ◽

Julio Saez-Rodriguez

Keyword(s):

Cell Line ◽

Cell Lines ◽

Protein Interactions ◽

Drug Response ◽

Prediction Models ◽

Response Prediction ◽

Single Type ◽

Disease Genes ◽

Cell Line Cell ◽

Network Diffusion

Download Full-text

Rapid proteotyping reveals cancer biology and drug response determinants in the NCI-60 cells

10.1101/268953 ◽

2018 ◽

Cited By ~ 3

Author(s):

Tiannan Guo ◽

Augustin Luna ◽

Vinodh N Rajapakse ◽

Ching Chiek Koh ◽

Zhicheng Wu ◽

...

Keyword(s):

Cell Lines ◽

Cancer Cell ◽

Cancer Biology ◽

Drug Response ◽

Missing Values ◽

Protein Complexes ◽

Cancer Cell Lines ◽

Molecular Data ◽

Data Matrix ◽

Proteomic Data

SummaryWe describe the rapid and reproducible acquisition of quantitative proteome maps for the NCI-60 cancer cell lines and their use to reveal cancer biology and drug response determinants. Proteome datasets for the 60 cell lines were acquired in duplicate within 30 working days using pressure cycling technology and SWATH mass spectrometry. We consistently quantified 3,171 SwissProt proteotypic proteins across all cell lines, generating a data matrix with 0.1% missing values, allowing analyses of protein complexes and pathway activities across all the cancer cells. Systematic and integrative analysis of the genetic variation, mRNA expression and proteomic data of the NCI-60 cancer cell lines uncovered complementarity between different types of molecular data in the prediction of the response to 240 drugs. We additionally identified novel proteomic drug response determinants for clinically relevant chemotherapeutic and targeted therapies. We anticipate that this study represents a landmark effort toward the translational application of proteotypes, which reveal biological insights that are easily missed in the absence of proteomic data.

Download Full-text

ADRML: anticancer drug response prediction using manifold learning

Scientific Reports ◽

10.1038/s41598-020-71257-7 ◽

2020 ◽

Vol 10 (1) ◽

Author(s):

Fatemeh Ahmadi Moughari ◽

Changiz Eslahchi

Keyword(s):

Cell Line ◽

Cell Lines ◽

Anticancer Drug ◽

Manifold Learning ◽

Drug Response ◽

Drug Information ◽

Response Prediction ◽

Drug Mechanism ◽

Line Drug ◽

Anticancer Drug Response

AbstractOne of the prominent challenges in precision medicine is to select the most appropriate treatment strategy for each patient based on the personalized information. The availability of massive data about drugs and cell lines facilitates the possibility of proposing efficient computational models for predicting anticancer drug response. In this study, we propose ADRML, a model for Anticancer Drug Response Prediction using Manifold Learning to systematically integrate the cell line information with the drug information to make accurate predictions about drug therapeutic. The proposed model maps the drug response matrix into the lower-rank spaces that lead to obtaining new perspectives about cell lines and drugs. The drug response for a new cell line-drug pair is computed using the low-rank features. The evaluation of ADRML performance on various types of cell lines and drug information, in addition to the comparisons with previously proposed methods, shows that ADRML provides accurate and robust predictions. Further investigations about the association between drug response and pathway activity scores reveal that the predicted drug responses can shed light on the underlying drug mechanism. Also, the case studies suggest that the predictions of ADRML about novel cell line-drug pairs are validated by reliable pieces of evidence from the literature. Consequently, the evaluations verify that ADRML can be used in accurately predicting and imputing the anticancer drug response.

Download Full-text

Looking at the BiG picture: Incorporating bipartite graphs in drug response prediction

10.1101/2021.08.11.455993 ◽

2021 ◽

Author(s):

David Earl Hostallero ◽

Yihui Li ◽

Amin Emad

Keyword(s):

Cell Line ◽

Bipartite Graph ◽

Cell Lines ◽

Drug Response ◽

Response Prediction ◽

Resistant Cell ◽

Resistant Cell Line ◽

Learning Approaches ◽

Convolutional Network ◽

Chemical Structures

Motivation: The increasing number of publicly available databases containing drugs' chemical structures, their response in cell lines, and molecular profiles of the cell lines has garnered attention to the problem of drug response prediction. However, many existing methods do not fully leverage the information that is shared among cell lines and drugs with similar structure. As such, drug similarities in terms of cell line responses and chemical structures could prove to be useful in forming drug representations to improve drug response prediction accuracy. Results: We present two deep learning approaches, BiG-DRP and BiG-DRP+, for drug response prediction. Our models take advantage of the drugs' chemical structure and the underlying relationships of drugs and cell lines through a bipartite graph and a heterogenous graph convolutional network that incorporate sensitive and resistant cell line information in forming drug representations. Evaluation of our methods and other state-of-the-art models in different scenarios show that incorporating this bipartite graph significantly improve the prediction performance. Additionally, genes that contribute significantly to the performance of our models also point to important biological processes and signaling pathways.

Download Full-text

Graph convolutional networks for drug response prediction

10.1101/2020.04.07.030908 ◽

2020 ◽

Author(s):

Tuan Nguyen ◽

Thin Nguyen ◽

Duc-Hau Le

Keyword(s):

Cell Line ◽

Cell Lines ◽

Drug Response ◽

Response Prediction ◽

Line Pair ◽

Convolutional Network ◽

Convolutional Networks ◽

Saliency Maps ◽

Genomic Aberrations ◽

Fully Connected

AbstractBackgroundDrug response prediction is an important problem in computational personalized medicine. Many machine learning-, especially deep learning-, based methods have been proposed for this task. However, these methods often represented the drugs as strings, which are not a natural way to depict molecules. Also, interpretation has not been considered thoroughly in these methods.MethodsIn this study, we propose a novel method, GraphDRP, based on graph convolutional network for the problem. In GraphDRP, drugs are represented in molecular graphs directly capturing the bonds among atoms, meanwhile cell lines are depicted as binary vectors of genomic aberrations. Representative features of drugs and cell lines are learned by convolution layers, then combined to represent for each drug-cell line pair. Finally, the response value of each drug-cell line pair is predicted by a fully-connected neural network. Four variants of graph convolutional networks are used for learning the features of drugs.ResultsWe find that GraphDRP outperforms tCNN in all performance measures for all experiments. Also, through saliency maps of the resulting GraphDRP models, we discover the contribution of the genomic aberrations to the responses.ConclusionRepresenting drugs as graphs are able to improve the performance of drug response prediction. Data and source code can be downloaded at https://github.com/hauldhut/GraphDRP.

Download Full-text

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

Genes ◽

10.3390/genes12060844 ◽

2021 ◽

Vol 12 (6) ◽

pp. 844

Author(s):

Abhishek Majumdar ◽

Yueze Liu ◽

Yaoqin Lu ◽

Shaofeng Wu ◽

Lijun Cheng

Keyword(s):

Gene Expression ◽

Support Vector Regression ◽

Cell Line ◽

Cell Lines ◽

Cancer Cell ◽

Drug Response ◽

Prediction Models ◽

Response Prediction ◽

Cancer Cell Lines ◽

Support Vector

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

Download Full-text

Cytotoxicity of chitosan ultrafine nanoshuttles on MCF-7 cell line as surrogate model for breast cancer

Current Drug Delivery ◽

10.2174/1567201817666200719005440 ◽

2020 ◽

Vol 17 ◽

Author(s):

Tarek Faris ◽

Gamaleldin I. Harisa ◽

Fars K. Alanazi ◽

Mohamed M. Badran ◽

Afraa Mohammad Alotaibi ◽

...

Keyword(s):

Red Blood Cells ◽

Factorial Design ◽

Cell Line ◽

Cell Lines ◽

Blood Cells ◽

Sodium Tripolyphosphate ◽

Physicochemical Characterization ◽

Spherical Shape ◽

Mcf 7

Aim: This study aimed to explore an affordable technique for the fabrication of Chitosan Nanoshuttles (CSNS) at the ultrafine nanoscale less than 100 nm with improved physicochemical properties, and cytotoxicity on the MCF-7 cell line. Background: Despite several studies reported that the antitumor effect of CS and CSNS could achieve intracellular compartment target ability, no enough available about this issue and further studies are required to address this assumption. Objectives: The objective of the current study was to investigate the potential processing variables for the production of ultrafine CSNS (> 100 nm) using Box-Benhken Design factorial design (BBD). This was achieved through a study of the effects of processing factors, such as CS concentration, CS/TPP ratio, and pH of the CS solution, on PS, PDI, and ZP. Moreover, the obtained CSNS was evaluated for physicochemical characteristics, morphology Also, hemocompatibility, and cytotoxicity using Red Blood Cells (RBCs) and MCF-7 cell lines were investigated. Methods: Box-Benhken Design factorial design (BBD) was used in the analysis of different selected variables. The effects of CS concentration, sodium tripolyphosphate (TPP) ratio, and pH on particle size, Polydispersity Index (PDI), and Zeta Potential (ZP) were measured. Subsequently, the prepared CS nanoshuttles were exposed to stability studies, physicochemical characterization, hemocompatibility, and cytotoxicity using red blood cells and MCF-7 cell lines as surrogate models for in vivo study. Result: The present results revealed that the optimized CSNS have ultrafine nanosize, (78.3±0.22 nm), homogenous with PDI (0.131±0.11), and ZP (31.9±0.25 mV). Moreover, CSNS have a spherical shape, amorphous in structure, and physically stable. Also, CSNS has biological safety as indicated by a gentle effect on red blood cell hemolysis, besides, the obtained nanoshuttles decrease MCF-7 viability. Conclusion: The present findings concluded that the developed ultrafine CSNS has unique properties with enhanced cytotoxicity. thus promising for use in intracellular organelles drug delivery.

Download Full-text

Transformation-associated alterations in interactions between pre-B cells and fibronectin

Blood ◽

10.1182/blood.v76.11.2311.2311 ◽

1990 ◽

Vol 76 (11) ◽

pp. 2311-2320 ◽

Cited By ~ 22

Author(s):

FM Lemoine ◽

S Dedhar ◽

GM Lima ◽

CJ Eaves

Keyword(s):

B Cells ◽

B Cell ◽

Cell Line ◽

Cell Lines ◽

Stromal Cells ◽

Stromal Cell ◽

Cell Attachment ◽

Attachment Site ◽

Receptor Antibodies

Abstract Marrow stromal elements produce as yet uncharacterized soluble growth factors that can stimulate the proliferation of murine pre-B cells, although close contact between these two cell types appears to ensure a better pre-B cell response. We have now shown that freshly isolated normal pre-B cells (ie, the B220+, surface mu- fraction of adult mouse bone marrow) adhere to fibronectin (FN) via an RGD cell-attachment site, as shown in a serum-free adherence assay, and they lose this functional ability on differentiation in vivo into B cells (ie, the B220+, surface mu+ fraction). Similarly, cells from an immortalized but stromal cell-dependent and nontumorigenic murine pre-B cell line originally derived from a Whitlock-Witte culture were also found to adhere to fibronectin (FN) via an RGD cell-attachment site. Moreover, in the presence of anti-FN receptor antibodies, the ability of this immortalized pre-B cell line to proliferate when co-cultured with a supportive stromal cell line (M2–10B4 cells) was markedly reduced (down to 30% of control). This suggests that pre-B cell attachment to FN on stromal cells may be an important component of the mechanism by which stromal cells stimulate normal pre-B cell proliferation and one that is no longer operative to control their more differentiated progeny. Two differently transformed pre-B cell lines, both of which are autocrine, stromal-independent, tumorigenic in vivo, and partially or completely differentiation-arrested at a very early stage of pre-B cell development, did not bind to FN. In addition, anti-FN receptor antibodies were much less effective in diminishing the ability of these tumorigenic pre-B cells to respond to M2–10B4 cell stimulation, which could still be demonstrated when the tumorigenic pre-B cells were co- cultured with M2–10B4 cells at a sufficiently low cell density. Analysis of cell surface molecules immunoprecipitated from both the nontumorigenic and tumorigenic pre-B cell lines by an anti-FN receptor antibody showed an increase in very late antigen (VLA) alpha chain(s) in both tumorigenic pre-B cell lines and a decrease in the beta 1 chain in one. Interestingly, all of the pre-B cell lines expressed similar amounts of messenger RNA for the beta 1 chain of the FN receptor. These results suggest that alteration of FN receptor expression on pre-B cells may represent a mechanism contributing to the outgrowth of leukemic pre-B cells with an autocrine phenotype and capable of stromal cell-independent, autonomous growth.

Download Full-text