Reconstructing cancer drug response networks using multitask learning

BackgroundTranslating in vitro results to clinical tests is a major challenge in systems biology. Here we present a new Multi-Task learning framework which integrates thousands of cell line expression experiments to reconstruct drug specific response networks in cancer.ResultsThe reconstructed networks correctly identify several shared key proteins and pathways while simultaneously highlighting many cell type specific proteins. We used top proteins from each drug network to predict survival for patients prescribed the drug.ConclusionsPredictions based on proteins from the in-vitro derived networks significantly outperformed predictions based on known cancer genes indicating that Multi-Task learning can indeed identify accurate drug response networks.

[1]  G. Bepler,et al.  Establishment and identification of small cell lung cancer cell lines having classic and variant features. , 1985, Cancer research.

[2]  M. Imamura,et al.  Characterization of 21 newly established esophageal cancer cell lines , 1992, Cancer.

[3]  Rich Caruana,et al.  Multitask Learning , 1998, Encyclopedia of Machine Learning and Data Mining.

[4]  C. L. La Porta,et al.  Overexpression of novel protein kinase C &dgr; in BL6 murine melanoma cells inhibits the proliferative capacity in vitro but enhances the metastatic potential in vivo , 2000, Melanoma research.

[5]  M. Ashburner,et al.  Gene Ontology: tool for the unification of biology , 2000, Nature Genetics.

[6]  Christian A. Rees,et al.  Systematic variation in gene expression patterns in human cancer cell lines , 2000, Nature Genetics.

[7]  E. Borden,et al.  IFN-β Pretreatment Sensitizes Human Melanoma Cells to TRAIL/Apo2 Ligand-Induced Apoptosis1 , 2002, The Journal of Immunology.

[8]  Jaana Kekäläinen,et al.  Cumulated gain-based evaluation of IR techniques , 2002, TOIS.

[9]  A. Pardee,et al.  Analysing differential gene expression in cancer , 2003, Nature Reviews Cancer.

[10]  H. Werner,et al.  Functional and Physical Interactions between BRCA1 and p53 in Transcriptional Regulation of the IGF-IR Gene , 2003, Hormone and metabolic research = Hormon- und Stoffwechselforschung = Hormones et metabolisme.

[11]  Jiawei Han,et al.  Cancer classification using gene expression data , 2003, Inf. Syst..

[12]  Audrey P. Gasch,et al.  The environmental stress response: a common yeast response to diverse environmental stresses , 2003 .

[13]  Paul T. Groth,et al.  The ENCODE (ENCyclopedia Of DNA Elements) Project , 2004, Science.

[14]  Tommi S. Jaakkola,et al.  Physical Network Models , 2004, J. Comput. Biol..

[15]  Charles A. Micchelli,et al.  Learning Multiple Tasks with Kernel Methods , 2005, J. Mach. Learn. Res..

[16]  Gregory N. Hullender,et al.  Learning to rank using gradient descent , 2005, ICML.

[17]  Pablo Tamayo,et al.  Gene set enrichment analysis: A knowledge-based approach for interpreting genome-wide expression profiles , 2005, Proceedings of the National Academy of Sciences of the United States of America.

[18]  P. Imming,et al.  Drugs, their targets and the nature and number of drug targets , 2006, Nature Reviews Drug Discovery.

[19]  Hongbing Shen,et al.  Variant genotypes of CDKN1A and CDKN1B are associated with an increased risk of breast cancer in Chinese women , 2006, International journal of cancer.

[20]  Trupti Joshi,et al.  Inferring gene regulatory networks from multiple microarray datasets , 2006, Bioinform..

[21]  J. Lamerdin,et al.  Identifying off-target effects and hidden phenotypes of drugs in human cells , 2006, Nature chemical biology.

[22]  N. Rahman,et al.  ATM and breast cancer susceptibility , 2006, Oncogene.

[23]  Wen-Lin Kuo,et al.  A collection of breast cancer cell lines for the study of functionally distinct cancer subtypes. , 2006, Cancer cell.

[24]  Marc Ladanyi,et al.  TLE1 as a Diagnostic Immunohistochemical Marker for Synovial Sarcoma Emerging From Gene Expression Profiling Studies , 2007, The American journal of surgical pathology.

[25]  W. Hong,et al.  Alien interacts with the human androgen receptor and inhibits prostate cancer cell growth. , 2007, Molecular endocrinology.

[26]  Massimiliano Pontil,et al.  Convex multi-task feature learning , 2008, Machine Learning.

[27]  Qizhi Yao,et al.  Profiling of 95 MicroRNAs in Pancreatic Cancer Cell Lines and Surgical Specimens by Real-Time PCR Analysis , 2009, World Journal of Surgery.

[28]  E. Xing,et al.  Statistical Estimation of Correlated Genome Associations to a Quantitative Trait Network , 2009, PLoS genetics.

[29]  E. Fraenkel,et al.  Integrating Proteomic, Transcriptional, and Interactome Data Reveals Hidden Components of Signaling and Regulatory Networks , 2009, Science Signaling.

[30]  P. Zapata,et al.  Tumoral Prostate Shows Different Expression Pattern of Somatostatin Receptor 2 (SSTR2) and Phosphotyrosine Phosphatase SHP-1 (PTPN6) According to Tumor Progression , 2009, Advances in urology.

[31]  Eric P. Xing,et al.  Tree-Guided Group Lasso for Multi-Task Regression with Structured Sparsity , 2009, ICML.

[32]  Paul Tseng,et al.  Trace Norm Regularization: Reformulations, Algorithms, and Multi-Task Learning , 2010, SIAM J. Optim..

[33]  Ernst-Walter Knapp,et al.  Exploring classification strategies with the CoEPrA 2006 contest , 2010, Bioinform..

[34]  Gunnar Rätsch,et al.  Leveraging Sequence Classification by Taxonomy-Based Multitask Learning , 2010, RECOMB.

[35]  Gregor Gößler,et al.  Efficient parameter search for qualitative models of regulatory networks using symbolic model checking , 2010, Bioinform..

[36]  Sheng Huang,et al.  TAL nucleases (TALNs): hybrid proteins composed of TAL effectors and FokI DNA-cleavage domain , 2010, Nucleic Acids Res..

[37]  N. Mazure,et al.  Constitutive ERK activity induces downregulation of tristetraprolin, a major protein controlling interleukin8/CXCL8 mRNA stability in melanoma cells. , 2011, American journal of physiology. Cell physiology.

[38]  A. Sivachenko,et al.  SF3B1 and other novel cancer genes in chronic lymphocytic leukemia. , 2011, The New England journal of medicine.

[39]  Hideaki Sugawara,et al.  DDBJ progress report , 2010, Nucleic Acids Res..

[40]  Anupam Gupta,et al.  Discovering pathways by orienting edges in protein interaction networks , 2010, Nucleic acids research.

[41]  S. Lipkowitz,et al.  WEE1 Inhibition Sensitizes Basal Breast Cancer Cells to TRAIL-Induced Apoptosis , 2011, Molecular Cancer Research.

[42]  Damian Szklarczyk,et al.  The STRING database in 2011: functional interaction networks of proteins, globally integrated and scored , 2010, Nucleic Acids Res..

[43]  Ernest Fraenkel,et al.  ResponseNet: revealing signaling and regulatory networks linking genetic and transcriptomic screening data , 2011, Nucleic Acids Res..

[44]  Mingming Jia,et al.  COSMIC: mining complete cancer genomes in the Catalogue of Somatic Mutations in Cancer , 2010, Nucleic Acids Res..

[45]  Yamir Moreno,et al.  Topological effects of data incompleteness of gene regulatory networks , 2012, BMC Systems Biology.

[46]  I. Simon,et al.  Studying and modelling dynamic biological processes using time-series gene expression data , 2012, Nature Reviews Genetics.

[47]  Nallasivam Palanisamy,et al.  Role of transcriptional corepressor CtBP1 in prostate cancer progression. , 2012, Neoplasia.

[48]  Zhi-Ren Liu,et al.  PDGF Upregulates Mcl-1 Through Activation of β-Catenin and HIF-1α-Dependent Signaling in Human Prostate Cancer Cells , 2012, PloS one.

[49]  Jieping Ye,et al.  Robust multi-task feature learning , 2012, KDD.

[50]  Jia Luo,et al.  Cross-talk between HER2 and MED1 regulates tamoxifen resistance of human breast cancer cells. , 2012, Cancer research.

[51]  Steven J. M. Jones,et al.  Comprehensive molecular portraits of human breast tumors , 2012, Nature.

[52]  J. Korbel,et al.  Genomic deletion of MAP3K7 at 6q12-22 is associated with early PSA recurrence in prostate cancer and absence of TMPRSS2:ERG fusions , 2013, Modern Pathology.

[53]  Ziv Bar-Joseph,et al.  Identifying proteins controlling key disease signaling pathways , 2013, Bioinform..

[54]  J. Bartlett,et al.  DNA repair and replication proteins as prognostic markers in melanoma , 2013, Histopathology.

[55]  Z. Bar-Joseph,et al.  Linking the signaling cascades and dynamic regulatory networks controlling stress responses , 2013, Genome research.

[56]  Steven J. M. Jones,et al.  Comprehensive molecular portraits of human breast tumours , 2013 .

[57]  S. Inoue,et al.  Amyloid precursor protein in human breast cancer: An androgen‐induced gene associated with cell proliferation , 2013, Cancer science.

[58]  Christie S. Chang,et al.  The BioGRID interaction database: 2013 update , 2012, Nucleic Acids Res..

[59]  Ziv Bar-Joseph,et al.  Predicting tissue specific transcription factor binding sites , 2013, BMC Genomics.

[60]  D. Haussler,et al.  The Somatic Genomic Landscape of Glioblastoma , 2013, Cell.

[61]  M. Fukata,et al.  A unique host defense pathway: TRIF mediates both antiviral and antibacterial immune responses. , 2013, Microbes and infection.

[62]  Joshua M. Stuart,et al.  The Cancer Genome Atlas Pan-Cancer analysis project , 2013, Nature Genetics.

[63]  Roded Sharan,et al.  iPoint: an integer programming based algorithm for inferring protein subnetworks. , 2013, Molecular bioSystems.

[64]  A. Ashworth,et al.  An siRNA screen identifies the GNAS locus as a driver in 20q amplified breast cancer , 2013, Oncogene.

[65]  M. Farach-Carson,et al.  IL‐1β Induces p62/SQSTM1 and Represses Androgen Receptor Expression in Prostate Cancer Cells , 2014, Journal of cellular biochemistry.

[66]  Christian Borgs,et al.  Sharing Information to Reconstruct Patient-Specific Pathways in Heterogeneous Diseases , 2013, Pacific Symposium on Biocomputing.

[67]  Ziv Bar-Joseph,et al.  Multitask Learning of Signaling and Regulatory Networks with Application to Studying Human Response to Flu , 2014, PLoS Comput. Biol..

[68]  M. Kwon,et al.  Expression of Histone Deacetylases HDAC1, HDAC2, HDAC3, and HDAC6 in Invasive Ductal Carcinomas of the Breast , 2014, Journal of breast cancer.

[69]  Andrew D. Rouillard,et al.  LINCS Canvas Browser: interactive web app to query, browse and interrogate LINCS L1000 gene expression signatures , 2014, Nucleic Acids Res..

[70]  Roded Sharan,et al.  Network orientation via shortest paths , 2014, Bioinform..

[71]  Scott D. Findlay,et al.  Assessing breast cancer cell lines as tumour models by comparison of mRNA expression profiles , 2015, Breast Cancer Research.

[72]  T. Gajewski,et al.  Innate immune recognition of cancer. , 2015, Annual review of immunology.

[73]  Yan Chen,et al.  Accuracy of GE digital breast tomosynthesis versus supplementary mammographic views for diagnosis of screen-detected soft tissue breast lesions , 2015, Breast Cancer Research.

[74]  Avi Ma'ayan,et al.  Drug-induced adverse events prediction with the LINCS L1000 data , 2016, Bioinform..

[75]  Hristo S. Paskov,et al.  Multitask learning improves prediction of cancer drug sensitivity , 2016, Scientific Reports.

[76]  Mehmet Tan.  Prediction of anti-cancer drug response by kernelized multi-task learning , 2016, Artif. Intell. Medicine.