scholarly journals Drug Sensitivity Prediction From Cell Line-Based Pharmacogenomics Data: Guidelines for Developing Machine Learning Models

2021 ◽  
Author(s):  
Hossein Sharifi-Noghabi ◽  
Soheil Jahangiri-Tazehkand ◽  
Casey Hon ◽  
Petr Smirnov ◽  
Anthony Mammoliti ◽  
...  
Keyword(s):  
Machine Learning ◽  
Cell Line ◽  
Cell Lines ◽  
Cancer Cell ◽  
Drug Sensitivity ◽  
Cancer Cell Lines ◽  
Precision Oncology ◽  
Learning Models ◽  
Drug Sensitivity Prediction ◽  
Machine Learning Models

ABSTRACTThe goal of precision oncology is to tailor treatment for patients individually using the genomic profile of their tumors. Pharmacogenomics datasets such as cancer cell lines are among the most valuable resources for drug sensitivity prediction, a crucial task of precision oncology. Machine learning methods have been employed to predict drug sensitivity based on the multiple omics data available for large panels of cancer cell lines. However, there are no comprehensive guidelines on how to properly train and validate such machine learning models for drug sensitivity prediction. In this paper, we introduce a set of guidelines for different aspects of training a predictor using cell line datasets. These guidelines provide extensive analysis of the generalization of drug sensitivity predictors, and challenge many current practices in the community including the choice of training dataset and measure of drug sensitivity. Application of these guidelines in future studies will enable the development of more robust preclinical biomarkers.

scholarly journals Interpretable deep recommender system model for prediction of kinase inhibitor efficacy across cancer cell lines

2021 ◽  
Author(s):  
Krzysztof Koras ◽  
Ewa Kizling ◽  
Dilafruz Juraeva ◽  
Eike Staub ◽  
Ewa Szczurek
Keyword(s):  
Neural Network ◽  
Cell Line ◽  
Cell Lines ◽  
Cancer Cell ◽  
Drug Sensitivity ◽  
Kinase Inhibitor ◽  
Cancer Cell Line ◽  
Cancer Cell Lines ◽  
Biological Processes ◽  
Drug Sensitivity Prediction

Computational models for drug sensitivity prediction have the potential to revolutionise personalized cancer medicine. Drug sensitivity assays, as well as profiling of cancer cell lines and drugs becomes increasingly available for training such models. Machine learning methods for drug sensitivity prediction must be optimized for: (i) leveraging the wealth of information about both cancer cell lines and drugs, (ii) predictive performance and (iii) interpretability. Multiple methods were proposed for predicting drug sensitivity from cancer cell line features, some in a multi-task fashion. So far, no such model leveraged drug inhibition profiles. Recent neural network-based recommender systems arise as models capable of predicting cancer cell line response to drugs from their biological features with high prediction accuracy. These models, however, require a tailored approach to model interpretability. In this work, we develop a neural network recommender system for kinase inhibitor sensitivity prediction called DEERS. The model utilizes molecular features of the cancer cell lines and kinase inhibition profiles of the drugs. DEERS incorporates two autoencoders to project cell line and drug features into 10-dimensional hidden representations and a feed-forward neural network to combine them into response prediction. We propose a novel model interpretability approach offering the widest possible assessment of the specific genes and biological processes that underlie the action of the drugs on the cell lines. The approach considers also such genes and processes that were not included in the set of modeled features. Our approach outperforms simpler matrix factorization models, achieving R=0.82 correlation between true and predicted response for the unseen cell lines. Using the interpretability analysis, we evaluate correlation of all human genes with each of the hidden cell line dimensions. Subsequently, we identify 67 biological processes associated with these dimensions. Combined with drug response data, these associations point at the processes that drive the cell line sensitivity to particular compounds. Detailed case studies are shown for PHA-793887, XMD14-99 and Dabrafenib. Our framework provides an expressive, multitask neural network model with a custom interpretability approach for inferring underlying biological factors and explaining cancer cell response to drugs.

scholarly journals Network-based clustering for drug sensitivity prediction in cancer cell lines

2019 ◽  
Author(s):  
Maryam Pouryahya ◽  
Jung Hun Oh ◽  
James C. Mathews ◽  
Zehor Belkhatir ◽  
Caroline Moosmüller ◽  
...  
Keyword(s):  
Cell Line ◽  
Cell Lines ◽  
Cancer Cell ◽  
Large Scale ◽  
Computational Models ◽  
Drug Sensitivity ◽  
Cancer Cell Lines ◽  
Machine Learning Techniques ◽  
Learning Techniques ◽  
Drug Sensitivity Prediction

AbstractThe study of large-scale pharmacogenomics provides an unprecedented opportunity to develop computational models that can accurately predict large cohorts of cell lines and drugs. In this work, we present a novel method for predicting drug sensitivity in cancer cell lines which considers both cell line genomic features and drug chemical features. Our network-based approach combines the theory of optimal mass transport (OMT) with machine learning techniques. It starts with unsupervised clustering of both cell line and drug data, followed by the prediction of drug sensitivity in the paired cluster of cell lines and drugs. We show that prior clustering of the heterogenous cell lines and structurally diverse drugs significantly improves the accuracy of the prediction. In addition, it facilities the interpretability of the results and identification of molecular biomarkers which are significant for both clustering of the cell lines and predicting the drug response.

scholarly journals Network-Based Matching of Patients and Targeted Therapies for Precision Oncology*

2019 ◽  
Author(s):  
Qingzhi Liu ◽  
Min Jin Ha ◽  
Rupam Bhattacharyya ◽  
Lana Garmire ◽  
Veerabhadran Baladandayuthapani
Keyword(s):  
Lung Cancer ◽  
Cancer Patients ◽  
Cell Lines ◽  
Cancer Cell ◽  
Drug Sensitivity ◽  
Cancer Cell Lines ◽  
Precision Oncology ◽  
Large Set ◽  
Sensitivity Data ◽  
The Right

The extensive acquisition of high-throughput molecular profiling data across model systems (human tumors and cancer cell lines) and drug sensitivity data, makes precision oncology possible – allowing clinicians to match the right drug to the right patient. Current supervised models for drug sensitivity prediction, often use cell lines as exemplars of patient tumors and for model training. However, these models are limited in their ability to accurately predict drug sensitivity of individual cancer patients to a large set of drugs, given the paucity of patient drug sensitivity data used for testing and high variability across different drugs. To address these challenges, we developed a multilayer network-based approach to impute individual patients’ responses to a large set of drugs. This approach considers the triplet of patients, cell lines and drugs as one inter-connected holistic system. We first use the omics profiles to construct a patient-cell line network and determine best matching cell lines for patient tumors based on robust measures of network similarity. Subsequently, these results are used to impute the “missing link” between each individual patient and each drug, called Personalized Imputed Drug Sensitivity Score (PIDS-Score), which can be construed as a measure of the therapeutic potential of a drug or therapy. We applied our method to two subtypes of lung cancer patients, matched these patients with cancer cell lines derived from 19 tissue types based on their functional proteomics profiles, and computed their PIDS-Scores to 251 drugs and experimental compounds. We identified the best representative cell lines that conserve lung cancer biology and molecular targets. The PIDS-Score based top sensitive drugs for the entire patient cohort as well as individual patients are highly related to lung cancer in terms of their targets, and their PIDS-Scores are significantly associated with patient clinical outcomes. These findings provide evidence that our method is useful to narrow the scope of possible effective patient-drug matchings for implementing evidence-based personalized medicine strategies.Data and code availabilityhttps://github.com/bayesrx/bayesrx.github.io/tree/master/authors/liu-q./ Shiny app (data and results visualization tool): https://qingzliu.shinyapps.io/psb-app/

Clinical drug response prediction from preclinical cancer cell lines by logistic matrix factorization approach

2021 ◽  
Author(s):  
Akram Emdadi ◽  
Changiz Eslahchi
Keyword(s):  
Cell Line ◽  
Cell Lines ◽  
Cancer Cell ◽  
Matrix Factorization ◽  
Drug Response ◽  
Drug Sensitivity ◽  
Cancer Cell Lines ◽  
Tissue Type ◽  
Factorization Approach ◽  
Clinical Drug

Predicting tumor drug response using cancer cell line drug response values for a large number of anti-cancer drugs is a significant challenge in personalized medicine. Predicting patient response to drugs from data obtained from preclinical models is made easier by the availability of different knowledge on cell lines and drugs. This paper proposes the TCLMF method, a predictive model for predicting drug response in tumor samples that was trained on preclinical samples and is based on the logistic matrix factorization approach. The TCLMF model is designed based on gene expression profiles, tissue type information, the chemical structure of drugs and drug sensitivity (IC 50) data from cancer cell lines. We use preclinical data from the Genomics of Drug Sensitivity in Cancer dataset (GDSC) to train the proposed drug response model, which we then use to predict drug sensitivity of samples from the Cancer Genome Atlas (TCGA) dataset. The TCLMF approach focuses on identifying successful features of cell lines and drugs in order to calculate the probability of the tumor samples being sensitive to drugs. The closest cell line neighbours for each tumor sample are calculated using a description of similarity between tumor samples and cell lines in this study. The drug response for a new tumor is then calculated by averaging the low-rank features obtained from its neighboring cell lines. We compare the results of the TCLMF model with the results of the previously proposed methods using two databases and two approaches to test the model’s performance. In the first approach, 12 drugs with enough known clinical drug response, considered in previous methods, are studied. For 7 drugs out of 12, the TCLMF can significantly distinguish between patients that are resistance to these drugs and the patients that are sensitive to them. These approaches are converted to classification models using a threshold in the second approach, and the results are compared. The results demonstrate that the TCLMF method provides accurate predictions across the results of the other algorithms. Finally, we accurately classify tumor tissue type using the latent vectors obtained from TCLMF’s logistic matrix factorization process. These findings demonstrate that the TCLMF approach produces effective latent vectors for tumor samples. The source code of the TCLMF method is available in https://github.com/emdadi/TCLMF.

scholarly journals Interpretable deep recommender system model for prediction of kinase inhibitor efficacy across cancer cell lines

2021 ◽  
Vol 11 (1) ◽  
Author(s):  
Krzysztof Koras ◽  
Ewa Kizling ◽  
Dilafruz Juraeva ◽  
Eike Staub ◽  
Ewa Szczurek
Keyword(s):  
Neural Network ◽  
Cell Line ◽  
Cell Lines ◽  
Cancer Cell ◽  
Recommender System ◽  
Computational Models ◽  
Drug Sensitivity ◽  
Kinase Inhibitor ◽  
Cancer Cell Lines ◽  
Molecular Features

AbstractComputational models for drug sensitivity prediction have the potential to significantly improve personalized cancer medicine. Drug sensitivity assays, combined with profiling of cancer cell lines and drugs become increasingly available for training such models. Multiple methods were proposed for predicting drug sensitivity from cancer cell line features, some in a multi-task fashion. So far, no such model leveraged drug inhibition profiles. Importantly, multi-task models require a tailored approach to model interpretability. In this work, we develop DEERS, a neural network recommender system for kinase inhibitor sensitivity prediction. The model utilizes molecular features of the cancer cell lines and kinase inhibition profiles of the drugs. DEERS incorporates two autoencoders to project cell line and drug features into 10-dimensional hidden representations and a feed-forward neural network to combine them into response prediction. We propose a novel interpretability approach, which in addition to the set of modeled features considers also the genes and processes outside of this set. Our approach outperforms simpler matrix factorization models, achieving R $$=$$ =  0.82 correlation between true and predicted response for the unseen cell lines. The interpretability analysis identifies 67 biological processes that drive the cell line sensitivity to particular compounds. Detailed case studies are shown for PHA-793887, XMD14-99 and Dabrafenib.

ANTITUMOR EFFECT OF SUNITINIB AND BORTEZOMIB ON BREAST CANCER CELL LINE IN VITRO

2017 ◽  
Vol 63 (1) ◽  
pp. 141-145
Author(s):  
Yuliya Khochenkova ◽  
Eliso Solomko ◽  
Oksana Ryabaya ◽  
Yevgeniya Stepanova ◽  
Dmitriy Khochenkov
Keyword(s):  
Breast Cancer ◽  
Cell Line ◽  
Cell Lines ◽  
Cancer Cell ◽  
Breast Cancer Cell ◽  
Antitumor Effect ◽  
Cancer Cell Lines ◽  
Breast Cancer Cell Lines ◽  
Actual Problem

The discovery for effective combinations of anticancer drugs for treatment for breast cancer is the actual problem in the experimental chemotherapy. In this paper we conducted a study of antitumor effect of the combination of sunitinib and bortezomib against MDA-MB-231 and SKBR-3 breast cancer cell lines in vitro. We found that bortezomib in non-toxic concentrations can potentiate the antitumor activity of sunitinib. MDA-MB-231 cell line has showed great sensitivity to the combination of bortezomib and sunitinib in vitro. Bortezomib and sunitinib caused reduced expression of receptor tyrosine kinases VEGFR1, VEGFR2, PDGFRa, PDGFRß and c-Kit on HER2- and HER2+ breast cancer cell lines

Synthesis and Characterization of Quinoline-3-Carboxamide Derivatives as Inhibitors of the ATM Kinase

2020 ◽  
Vol 20 (23) ◽  
pp. 2070-2079
Author(s):  
Srimadhavi Ravi ◽  
Sugata Barui ◽  
Sivapriya Kirubakaran ◽  
Parul Duhan ◽  
Kaushik Bhowmik
Keyword(s):  
Western Blot ◽  
Cancer Cells ◽  
Cell Line ◽  
Cell Lines ◽  
Cancer Cell ◽  
Blot Analysis ◽  
Western Blot Analysis ◽  
Cancer Cell Lines ◽  
Atm Kinase ◽  
Cytotoxicity Studies

Background: The importance of inhibiting the kinases of the DDR pathway for radiosensitizing cancer cells is well established. Cancer cells exploit these kinases for their survival, which leads to the development of resistance towards DNA damaging therapeutics. Objective: In this article, the focus is on targeting the key mediator of the DDR pathway, the ATM kinase. A new set of quinoline-3-carboxamides, as potential inhibitors of ATM, is reported. Methods: Quinoline-3-carboxamide derivatives were synthesized and cytotoxicity assay was performed to analyze the effect of molecules on different cancer cell lines like HCT116, MDA-MB-468, and MDA-MB-231. Results: Three of the synthesized compounds showed promising cytotoxicity towards a selected set of cancer cell lines. Western Blot analysis was also performed by pre-treating the cells with quercetin, a known ATM upregulator, by causing DNA double-strand breaks. SAR studies suggested the importance of the electron-donating nature of the R group for the molecule to be toxic. Finally, Western-Blot analysis confirmed the down-regulation of ATM in the cells. Additionally, the PTEN negative cell line, MDA-MB-468, was more sensitive towards the compounds in comparison with the PTEN positive cell line, MDA-MB-231. Cytotoxicity studies against 293T cells showed that the compounds were at least three times less toxic when compared with HCT116. Conclusion: In conclusion, these experiments will lay the groundwork for the evolution of potent and selective ATM inhibitors for the radio- and chemo-sensitization of cancer cells.

Synthesis, Anticancer Activity on Prostate Cancer Cell Lines and Molecular Modeling Studies of Flurbiprofen-Thioether Derivatives as Potential Target of MetAP (Type II)

2020 ◽  
Vol 16 (6) ◽  
pp. 735-749 ◽  
Author(s):  
Özgür Yılmaz ◽  
Burak Bayer ◽  
Hatice Bekçi ◽  
Abdullahi I. Uba ◽  
Ahmet Cumaoğlu ◽  
...  
Keyword(s):  
Prostate Cancer ◽  
Molecular Modeling ◽  
Cell Line ◽  
Cell Lines ◽  
Cancer Cell ◽  
Prostate Cancer Cell ◽  
Cancer Cell Line ◽  
Cancer Cell Lines ◽  
Prostate Cancer Cell Lines ◽  
Molecular Modeling Studies

Background:: Prostate cancer is still one of the serious causes of mortality and morbidity in men. Despite recent advances in anticancer therapy, there is a still need of novel agents with more efficacy and specificity in the treatment of prostate cancer. Because of its function on angiogenesis and overexpression in the prostate cancer, methionine aminopeptidase-2 (MetAP-2) has been a potential target for novel drug design recently. Objective:: A novel series of Flurbiprofen derivatives N-(substituted)-2-(2-(2-fluoro-[1,1'- biphenyl]-4-il)propanoyl)hydrazinocarbothioamide (3a-c), 4-substituted-3-(1-(2-fluoro-[1,1'-biphenyl]- 4-yl)ethyl)-1H-1,2,4-triazole-5(4H)-thione (4a-d), 3-(substitutedthio)-4-(substituted-phenyl)- 5-(1-(2-fluoro-[1,1'-biphenyl]-4-yl)ethyl)-4H-1,2,4-triazole (5a-y) were synthesized. The purpose of the research was to evaluate these derivatives against MetAP-2 in vitro and in silico to obtain novel specific and effective anticancer agents against prostate cancer. Methods: The chemical structures and purities of the compounds were defined by spectral methods (1H-NMR, 13C-NMR, HR-MS and FT-IR) and elemental analysis. Anticancer activities of the compounds were evaluated in vitro by using MTS method against PC-3 and DU-143 (androgenindependent human prostate cancer cell lines) and LNCaP (androgen-sensitive human prostate adenocarcinoma) prostate cancer cell lines. Cisplatin was used as a positive sensitivity reference standard. Results:: Compounds 5b and 5u; 3c, 5b and 5y; 4d and 5o showed the most potent biological activity against PC3 cancer cell line (IC50= 27.1 μM, and 5.12 μM, respectively), DU-145 cancer cell line (IC50= 11.55 μM, 6.9 μM and 9.54 μM, respectively) and LNCaP cancer cell line (IC50= 11.45 μM and 26.91 μM, respectively). Some compounds were evaluated for their apoptotic caspases protein expression (EGFR/PI3K/AKT pathway) by Western blot analysis in androgen independent- PC3 cells. BAX, caspase 9, caspsase 3 and anti-apoptotic BcL-2 mRNA levels of some compounds were also investigated. In addition, molecular modeling studies of the compounds on MetAP-2 enzyme active site were evaluated in order to get insight into binding mode and energy. Conclusion:: A series of Flurbiprofen-thioether derivatives were synthesized. This study presented that some of the synthesized compounds have remarkable anticancer and apoptotic activities against prostate cancer cells. Also, molecular modeling studies exhibited that there is a correlation between molecular modeling and anticancer activity results.

Sign in / Sign up

Export Citation Format

Share Document