Gene expression has more power for predicting in vitro cancer cell vulnerabilities than genomics

Achieving precision oncology requires accurate identification of targetable cancer vulnerabilities in patients. Generally, genomic features are regarded as the state-of-the-art method for stratifying patients for targeted therapies. In this work, we conduct the first rigorous comparison of DNA- and expression-based predictive models for viability across five datasets encompassing chemical and genetic perturbations. We find that expression consistently outperforms DNA for predicting vulnerabilities, including many currently stratified by canonical DNA markers. Contrary to their perception in the literature, the most accurate expression-based models depend on few features and are amenable to biological interpretation. This work points to the importance of exploring more comprehensive expression profiling in clinical settings.

[1]  Debashis Ghosh,et al.  Integrating Clinical and Multiple Omics Data for Prognostic Assessment across Human Cancers , 2017, Scientific Reports.

[2]  P. Chène.  Inhibiting the p53–MDM2 interaction: an important target for cancer therapy , 2003, Nature Reviews Cancer.

[3]  N. Blüthgen,et al.  The cancer cell proteome and transcriptome predicts sensitivity to targeted and cytotoxic drugs , 2019, Life Science Alliance.

[4]  Laura M. Heiser,et al.  A community effort to assess and improve drug sensitivity prediction algorithms , 2014, Nature Biotechnology.

[5]  A. Letai,et al.  Functional precision cancer medicine—moving beyond pure genomics , 2017, Nature Medicine.

[6]  Evan O. Paull,et al.  A Modular Master Regulator Landscape Determines the Impact of Genetic Alterations on the Transcriptional Identity of Cancer Cells , 2019, bioRxiv.

[7]  J Jack Lee,et al.  Genomic and transcriptomic profiling expands precision cancer medicine: the WINTHER trial , 2019, Nature Medicine.

[8]  Vinay Prasad,et al.  Perspective: The precision-oncology illusion , 2016, Nature.

[9]  P. Chène.  INHIBITING THE p 53 – MDM 2 INTERACTION : AN IMPORTANT TARGET FOR CANCER THERAPY , 2022 .

[10]  Jin Gu,et al.  Evaluating the molecule-based prediction of clinical drug responses in cancer , 2016, Bioinform..

[11]  N. Cox,et al.  Clinical drug response can be predicted using baseline gene expression levels and in vitro drug sensitivity in cell lines , 2014, Genome Biology.

[12]  Emanuel J. V. Gonçalves,et al.  A Landscape of Pharmacogenomic Interactions in Cancer , 2016, Cell.

[13]  Helga Thorvaldsdóttir,et al.  Molecular signatures database (MSigDB) 3.0 , 2011, Bioinform..

[14]  Y. Saeys,et al.  Radiation-induced alternative transcription and splicing events and their applicability to practical biodosimetry , 2016, Scientific Reports.

[15]  Aviad Tsherniak,et al.  Mutational processes shape the landscape of TP53 mutations in human cancer , 2018, Nature Genetics.

[16]  En-guo Chen,et al.  Nkx2-1: a novel tumor biomarker of lung cancer , 2012, Journal of Zhejiang University SCIENCE B.

[17]  R. Warters,et al.  Differential Gene Expression in Primary Human Skin Keratinocytes and Fibroblasts in Response to Ionizing Radiation , 2009, Radiation research.

[18]  Moriah H Nissan,et al.  OncoKB: A Precision Oncology Knowledge Base. , 2017, JCO precision oncology.

[19]  Xiaoling Li,et al.  Predicting tumor response to drugs based on gene-expression biomarkers of sensitivity learned from cancer cell lines , 2020, BMC Genomics.

[20]  L. Garraway,et al.  Clinical implications of the cancer genome. , 2010, Journal of clinical oncology : official journal of the American Society of Clinical Oncology.

[21]  G. Getz,et al.  Predicting and affecting response to cancer therapy based on pathway-level biomarkers , 2020, Nature Communications.

[22]  S. Brunak,et al.  A scored human protein–protein interaction network to catalyze genomic interpretation , 2017, Nature Methods.

[23]  Andreas Ruepp,et al.  CORUM: the comprehensive resource of mammalian protein complexes—2019 , 2018, Nucleic Acids Res..

[24]  Tero Aittokallio,et al.  Machine learning and feature selection for drug response prediction in precision oncology applications , 2018, Biophysical Reviews.

[25]  Lodewyk F. A. Wessels,et al.  TANDEM: a two-stage approach to maximize interpretability of drug response models based on multiple molecular data types , 2016, Bioinform..

[26]  Astrid Gall,et al.  Ensembl 2018 , 2017, Nucleic Acids Res..

[27]  K. Flaherty,et al.  Toward Minimal Residual Disease-Directed Therapy in Melanoma , 2018, Cell.

[28]  M. Berger,et al.  Clinical tumour sequencing for precision oncology: time for a universal strategy , 2018, Nature Reviews Cancer.

[29]  Andrei Yakovlev,et al.  Diverse correlation structures in gene expression data and their utility in improving statistical inference , 2007, 0712.2130.

[30]  Marcela Franco,et al.  Prediction of response to anti-cancer drugs becomes robust via network integration of molecular data , 2019, Scientific Reports.

[31]  Mariano J. Alvarez,et al.  Network-based inference of protein activity helps functionalize the genetic landscape of cancer , 2016, Nature Genetics.

[32]  Aviad Tsherniak,et al.  Extracting Biological Insights from the Project Achilles Genome-Scale CRISPR Screens in Cancer Cell Lines , 2019, bioRxiv.

[33]  Hans-Werner Mewes,et al.  CORUM: the comprehensive resource of mammalian protein complexes , 2007, Nucleic Acids Res..

[34]  Aviad Tsherniak,et al.  Improved estimation of cancer dependencies from large-scale RNAi screens using model-based normalization and data integration , 2018, Nature Communications.

[35]  J. Marine,et al.  Melanoma plasticity and phenotypic diversity: therapeutic barriers and opportunities , 2019, Genes & development.

[36]  Michael T. Zimmermann,et al.  Predict drug sensitivity of cancer cells with pathway activity inference , 2019, BMC Medical Genomics.

[37]  Nci Dream Community.  A community effort to assess and improve drug sensitivity prediction algorithms , 2014 .

[38]  Emanuel J. V. Gonçalves,et al.  Prioritization of cancer therapeutic targets using CRISPR–Cas9 screens , 2019, Nature.

[39]  Antoine de Weck,et al.  Project DRIVE: A Compendium of Cancer Dependencies and Synthetic Lethal Relationships Uncovered by Large-Scale, Deep RNAi Screening , 2017, Cell.

[40]  Joshua A. Bittker,et al.  Non-oncology drugs are a source of previously unappreciated anti-cancer activity , 2019, bioRxiv.

[41]  J. Baselga,et al.  The Evolving War on Cancer , 2011, Cell.

[42]  Jun Lu,et al.  Transcriptome profiling analysis reveals that CXCL2 is involved in anlotinib resistance in human lung cancer cells , 2019, BMC Medical Genomics.

[43]  W. Reinhold,et al.  Putative DNA/RNA helicase Schlafen-11 (SLFN11) sensitizes cancer cells to DNA-damaging agents , 2012, Proceedings of the National Academy of Sciences.