A Landscape of Pharmacogenomic Interactions in Cancer.

Iorio F, Knijnenburg TA, Vis DJ, Bignell GR, Menden MP, Schubert M, Aben N, Gonçalves E, Barthorpe S, Lightfoot H, Cokelaer T, Greninger P, van Dyk E, Chang H, de Silva H, Heyn H, Deng X, Egan RK, Liu Q, Mironenko T, Mitropoulos X, Richardson L, Wang J, Zhang T, Moran S, Sayols S, Soleimani M, Tamborero D, Lopez-Bigas N, Ross-Macdonald P, Esteller M, Gray NS, Haber DA, Stratton MR, Benes CH, Wessels LFA, Saez-Rodriguez J, McDermott U, Garnett MJ. · 2016

Read A Landscape of Pharmacogenomic Interactions in Cancer. by Iorio F, Knijnenburg TA, Vis DJ, Bignell GR, Menden MP, Schubert M, Aben N, Gonçalves E, Barthorpe S, Lightfoot H, Cokelaer T, Greninger P, van Dyk E, Chang H, de Silva H, Heyn H, Deng X, Egan RK, Liu Q, Mironenko T, Mitropoulos X, Richardson L, Wang J, Zhang T, Moran S, Sayols S, Soleimani M, Tamborero D, Lopez-Bigas N, Ross-Macdonald P, Esteller M, Gray NS, Haber DA, Stratton MR, Benes CH, Wessels LFA, Saez-Rodriguez J, McDermott U, Garnett MJ. (2016) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/europmc-pmc4967469.txt.

Introduction

Francesco Iorio leads a large team mapping how cancer gene changes predict drug response. Their study appeared in Cell in 2016. Tumor sequencing reveals each cancer's mutations, but turning that into a drug choice is hard. The team gathered alterations from more than eleven thousand tumors across many tissues. They mapped them onto one thousand and one human cancer cell lines, then tested those lines against hundreds of drugs.

Cell lines faithfully echoed the driver changes seen in real tumors. Many changes tracked with sensitivity or resistance to specific drugs. Tissue of origin also shaped response, sometimes as much as mutation did. Logic models found alteration pairs that sensitize cells, and machine learning ranked which data types predict response best. The authors offer the datasets as a public resource for matching tumor subtypes to treatments.

Catalog record

CREATORIorio F, Knijnenburg TA, Vis DJ, Bignell GR, Menden MP, Schubert M, Aben N, Gonçalves E, Barthorpe S, Lightfoot H, Cokelaer T, Greninger P, van Dyk E, Chang H, de Silva H, Heyn H, Deng X, Egan RK, Liu Q, Mironenko T, Mitropoulos X, Richardson L, Wang J, Zhang T, Moran S, Sayols S, Soleimani M, Tamborero D, Lopez-Bigas N, Ross-Macdonald P, Esteller M, Gray NS, Haber DA, Stratton MR, Benes CH, Wessels LFA, Saez-Rodriguez J, McDermott U, Garnett MJ.
YEAR2016
COLLECTIONPAPERS
FORMATTXT
FILE SIZE81 KB
LICENSECC BY 4.0
RIGHTSCC BY
SOURCEEurope PMC

Source: https://europepmc.org/articles/PMC4967469

Browse the full library · About the Tower of Babel

Full text

The opening of the work, from the library's hosted text:

A Landscape of Pharmacogenomic Interactions in Cancer By Francesco Iorio, Theo A Knijnenburg, Daniel J Vis, Graham R Bignell, Michael P Menden, Michael Schubert, Nanne Aben, Emanuel Gonçalves, Syd Barthorpe, Howard Lightfoot, Thomas Cokelaer, Patricia Greninger Published in: Cell DOI: 10.1016/j.cell.2016.06.017 ABSTRACT Summary Systematic studies of cancer genomes have provided unprecedented insights into the molecular nature of cancer. Using this information to guide the development and application of therapies in the clinic is challenging. Here, we report how cancer-driven alterations identified in 11,289 tumors from 29 tissues (integrating somatic mutations, copy number alterations, DNA methylation, and gene expression) can be mapped onto 1,001 molecularly annotated human cancer cell lines and correlated with sensitivity to 265 drugs. We find that cell lines faithfully recapitulate oncogenic alterations identified in tumors, find that many of these associate with drug sensitivity/resistance, and highlight the importance of tissue lineage in mediating drug response. Logic-based modeling uncovers combinations of alterations that sensitize to drugs, while machine learning demonstrates the relative importance of different data types in predicting drug response. Our analysis and datasets are rich resources to link genotypes with cellular phenotypes and to identify therapeutic options for selected cancer sub-populations. INTRODUCTION Cancers arise because of the acquisition of somatic alterations in their genomes that alter the function of key cancer genes (Stratton et al., 2009). A number of these alterations are implicated as determinants of treatment response in the clinic (Chapman et al., 2011, Mok et al., 2009, Shaw et al., 2013). Studies from The Cancer Genome Atlas (TCGA) and the International Cancer Genome Consortium (ICGC) have generated comprehensive catalogs of the cancer genes involved in tumorigenesis across a broad range of cancer types (Lawrence et al., 2014, Tamborero et al., 2013b, Zack et al., 2013). The emerging landscape of oncogenic alterations in cancer points to a hierarchy of likely functional processes and pathways that may guide the future treatment of patients (Ciriello et al., 2013, Hanahan and Weinberg, 2000, Stratton et al., 2009). Clinical trials are complex and expensive, and pre-clinical data that helps stratify patients can dramatically increase the likelihood of success during clinical development (Cook et al., 2014, Nelson et al., 2015). Thus, pre-clinical biological models that, as much as reasonably possible, capture both the molecular features of cancer and the diversity of therapeutic responses are a necessity. Human cancer cell lines are a facile experimental model and are widely used for drug development. Large-scale drug sensitivity screens in cancer cell lines have been used to identify clinically meaningful gene-drug interactions (Barretina et al., 2012, Basu et al., 2013, Garnett et al., 2012, Seashore-Ludlow et al., 2015). In the past, such screens have labored under the limitation of an imperfect understanding of the landscape of cancer driver genes, but it is now possible to view drug sensitivity in such models through the lens of clinically relevant oncogenic alterations. Here, we analyzed somatic mutations, copy number alterations, and hypermethylation across a total of 11,289 tumor samples from 29 tumor types to define a clinically relevant catalog of recurrent mutated cancer genes, focal amplifications/deletions, and methylated gene promoters (Figure 1A; Tables S1A–S1D). These oncogenic alterations were investigated as possible predictors of differential drug sensitivity across 1,001 cancer cell lines (Figures 1B and 1C; Table S1E) screened with 265 anti-cancer compounds (Figures 1D and S1; Table S1F). We have carried out an exploration of these data to determine (1) the extent to which cancer cell lines recapitulate oncogenic alterations in primary tumors, (2) which oncogenic alterations associate with drug sensitivity, (3) whether logic combinations of multiple alterations better explain drug sensitivity, and (4) the relative contribution of different molecular data types, either individually or in combination, in predicting drug response (Figure 1E). [Figure: Overview of Data and Analyses(A) Publicly available genomic data for a large cohort of primary tumors were analyzed to identify clinically relevant features called cancer functional events.(B) A panel of 1,001 genomically characterized human cancer cell lines.(C) The catalog of CFEs from patient tumors was used to filter the set of molecular alterations identified in cell lines and subsequently was used for pharmacogenomic modeling.(D) Cancer cell lines were screened for differential sensitivity against 265 anti-cancer compounds.(E) The resultant datasets were used for pharmacogenomic modeling.See also Figure S1 and Table S1.] [Figure: Screened Compound Duplicates, Related to Figure 1Histograms, scatter plots and Pearson correlation scores between IC50 profiles for 7 compounds screened in biological duplicates. In all cases replicate data were generated at least one year apart. Superimposed to each scatter plot is a contingency table (and a corresponding Fisher exact test p-value) showing consistency of sensitive (IC50 ≤ maximal tested concentration) and resistant (IC50 > maximal tested concentration) cell lines across replicates.] RESULTS Oncogenic Alterations in Human Tumors We built a comprehensive map of the oncogenic alterations in human tumors using data from TCGA, ICGC, and other studies (Figure 1A; Table S1C). The map consisted of (1) cancer genes (CGs) for which the mutation pattern in whole-exome sequencing (WES) data is consistent with positive selection, 2) focal recurrently aberrant copy number segments (RACSs) from SNP6 array profiles, and 3) hypermethylated informative 5′C-phosphate-G-3′ sites in gene promoters (iCpGs) from DNA methylation data, hereafter collectively referred to as “Cancer functional events” (CFEs). We identified CFEs by combining data across all tumors (pan-cancer), as well as for each cancer type (cancer specific) (Tables S2A, S2D, and S2H). The WES dataset consisted of somatic variant calls from 48 studies of matched tumor-normal samples, comprising 6,815 samples and spanning 28 cancer types (Tables S1A–S1D). CGs were detected per cancer type by combining the outputs of three algorithms: MutSigCV, OncodriveFM, and OncodriveCLUST (Lawrence et al., 2013, Rubio-Perez et al., 2015, Tamborero et al., 2013a). This identified 461 unique pan-cancer genes (Table S2A). We further added nine genes identified as putative tumor suppressors (Wong et al., 2014). We mined the COSMIC database to identify likely driver mutations in 358 of the 470 CGs (Table S2B; Supplemental Experimental Procedures). Most tumors harbored only a few driver mutations (median n = 2, range 0–64), consistent with previous reports (Kandoth et al., 2013, Vogelstein et al., 2013). RACSs were identified using ADMIRE for the analysis of 8,239 copy number arrays spanning 27 cancer types (van Dyk et al., 2013) (Table S1D; Supplemental Experimental Procedures). In total, 851 cancer-specific RACSs were gained (286 segments) or lost (565 segments), with a median of 19 RACSs per tumor type (Table S2D). The median number of genes within each RACS was 15 for amplified regions and one for deleted regions. The majority of known driver gene amplifications (e.g., EGFR, ERBB2, MET, and MYC) and homozygous deletions (e.g., CDKN2A, PTEN, and RB1) were captured, with 320 RACSs (38%) containing at least one known putative cancer driver gene, in addition to 531 RACSs (62%) without known driver genes. A smaller pan-cancer set (due to overlap in RACSs across cancer types) was constructed by pooling these results, comprising 425 RACSs (117 amplified and 308 deleted) (Tables S2D–S2F). iCpGs were identified using DNA methylation array data for 6,166 tumor samples spanning 21 cancer types (Table S1D). We defined 378 iCpGs based on a multimodal distribution of their methylation signal in at least one cancer type (Tables S2H and S2I). This

Read the complete text