Integrated deep learned transcriptomic and structure-based predictor of clinical trials outcomes

Despite many recent advances in systems biology and a marked increase in the availability of high-throughput biological data, the productivity of research and development in the pharmaceutical industry is on the decline. This is primarily due to clinical trial failure rates reaching up to 95% in oncology and other disease areas. We have developed a comprehensive analytical and computational pipeline utilizing deep learning techniques and novel systems biology analytical tools to predict the outcomes of phase I/II clinical trials. The pipeline predicts the side effects of a drug using deep neural networks and estimates drug-induced pathway activation. It then uses the predicted side effect probabilities and pathway activation scores as an input to train a classifier which predicts clinical trial outcomes. This classifier was trained on 577 transcriptomic datasets and has achieved a cross-validated accuracy of 0.83. When compared to a direct gene-based classifier, our multi-stage approach dramatically improves the accuracy of the predictions. The classifier was applied to a set of compounds currently present in the pipelines of several major pharmaceutical companies to highlight potential risks in their portfolios and estimate the fraction of clinical trials that were likely to fail in phase I and II.

[1]  Stephen R. Johnson,et al.  Molecular properties that influence the oral bioavailability of drug candidates. , 2002, Journal of medicinal chemistry.

[2]  Kenneth H. Buetow,et al.  PID: the Pathway Interaction Database , 2008, Nucleic Acids Res..

[3]  R. W. Hansen,et al.  Journal of Health Economics , 2016 .

[4]  G. Clermont,et al.  In silico design of clinical trials: A method coming of age , 2004, Critical care medicine.

[5]  Polina Mamoshina,et al.  Design of efficient computational workflows for in silico drug repurposing. , 2017, Drug discovery today.

[6]  Michael Hay,et al.  Clinical development success rates for investigational drugs , 2014, Nature Biotechnology.

[7]  C. Shapiro,et al.  Randomized phase II trial of the cyclin-dependent kinase inhibitor dinaciclib (MK-7965) versus capecitabine in patients with advanced breast cancer. , 2014, Clinical breast cancer.

[8]  D. Mann,et al.  Rethinking Phase II Clinical Trial Design in Heart Failure. , 2013, Clinical investigation.

[9]  G. V. Paolini,et al.  Quantifying the chemical beauty of drugs. , 2012, Nature chemistry.

[10]  Navdeep Jaitly,et al.  Adversarial Autoencoders , 2015, ArXiv.

[11]  Bernard H. Munos,et al.  How to Revive Breakthrough Innovation in the Pharmaceutical Industry , 2011, Science Translational Medicine.

[12]  Gilles Clermont,et al.  A Patient-Specific in silico Model of Inflammation and Healing Tested in Acute Vocal Fold Injury , 2008, PloS one.

[13]  Nina Jeliazkova,et al.  AMBIT RESTful web services: an implementation of the OpenTox application programming interface , 2011, J. Cheminformatics.

[14]  Heidi Ledford,et al.  Translational research: 4 ways to fix the clinical trial , 2011, Nature.

[15]  Yoram Singer,et al.  Adaptive Subgradient Methods for Online Learning and Stochastic Optimization , 2011, J. Mach. Learn. Res..

[16]  F. Lombardo,et al.  Experimental and computational approaches to estimate solubility and permeability in drug discovery and development settings. , 2001, Advanced drug delivery reviews.

[17]  Yoshua Bengio,et al.  Generative Adversarial Nets , 2014, NIPS.

[18]  Luigi Ferrucci,et al.  Deep biomarkers of aging are population-dependent , 2016, Aging.

[19]  Hiroyuki Ogata,et al.  KEGG: Kyoto Encyclopedia of Genes and Genomes , 1999, Nucleic Acids Res..

[20]  Yohan Payan,et al.  Patient specific finite element model of the face soft tissues for computer-assisted maxillofacial surgery , 2003, Medical Image Anal..

[21]  E. Brown,et al.  The Medical Dictionary for Regulatory Activities (MedDRA) , 1999, Drug safety.

[22]  Peer Bork,et al.  The SIDER database of drugs and side effects , 2015, Nucleic Acids Res..

[23]  Grace Patlewicz,et al.  Current and Future Perspectives on the Development, Evaluation, and Application of in Silico Approaches for Predicting Toxicity. , 2016, Chemical research in toxicology.

[24]  Ying Luan,et al.  A Phase I/II Multicenter, Open-Label Study of the Oral Histone Deacetylase Inhibitor Abexinostat in Relapsed/Refractory Lymphoma , 2015, Clinical Cancer Research.

[25]  Paul A Clemons,et al.  The Connectivity Map: Using Gene-Expression Signatures to Connect Small Molecules, Genes, and Disease , 2006, Science.

[26]  Kaitlyn M. Gayvert,et al.  A Data-Driven Approach to Predicting Successes and Failures of Clinical Trials. , 2016, Cell chemical biology.

[27]  Ben Goldacre,et al.  The TrialsTracker: Automated ongoing monitoring of failure to share clinical trial results by all major companies and research institutions , 2016, F1000Research.

[28]  J. Scannell,et al.  Diagnosing the decline in pharmaceutical R&D efficiency , 2012, Nature Reviews Drug Discovery.

[29]  Marco Viceconti,et al.  In silico clinical trials: how computer simulation will transform the biomedical industry , 2016 .

[30]  Alex Zhavoronkov,et al.  Applications of Deep Learning in Biomedicine. , 2016, Molecular pharmaceutics.

[31]  P J Hunter,et al.  An anatomically based patient-specific finite element model of patella articulation: towards a diagnostic tool , 2005, Biomechanics and modeling in mechanobiology.

[32]  T. Ashburn,et al.  Drug repositioning: identifying and developing new uses for existing drugs , 2004, Nature Reviews Drug Discovery.

[33]  D. Gary Gilliland,et al.  Drug development and clinical trials—the path to an approved cancer drug , 2012, Nature Reviews Clinical Oncology.

[34]  Guigang Zhang,et al.  Deep Learning , 2016, Int. J. Semantic Comput..

[35]  Matthew D. Zeiler.  ADADELTA: An Adaptive Learning Rate Method , 2012, ArXiv.

[36]  Jimmy Ba,et al.  Adam: A Method for Stochastic Optimization , 2014, ICLR.

[37]  Sergey Plis,et al.  Deep Learning Applications for Predicting Pharmacological Properties of Drugs and Drug Repurposing Using Transcriptomic Data. , 2016, Molecular pharmaceutics.

[38]  Nitish Srivastava,et al.  Dropout: a simple way to prevent neural networks from overfitting , 2014, J. Mach. Learn. Res..

[39]  Andrey Kazennov,et al.  The cornucopia of meaningful leads: Applying deep adversarial autoencoders for new molecule development in oncology , 2016, Oncotarget.

[40]  A. Aliper,et al.  In silico Pathway Activation Network Decomposition Analysis (iPANDA) as a method for biomarker development , 2016, Nature Communications.

[41]  J. Nemunaitis,et al.  Randomized phase 2 study of the cyclin-dependent kinase inhibitor dinaciclib (MK-7965) versus erlotinib in patients with non-small cell lung cancer. , 2014, Lung cancer.

[42]  A. Ghose,et al.  A knowledge-based approach in designing combinatorial or medicinal chemistry libraries for drug discovery. 1. A qualitative and quantitative characterization of known drug databases. , 1999, Journal of combinatorial chemistry.

[43]  João Rodrigues,et al.  Why and how have drug discovery strategies in pharma changed? What are the new mindsets? , 2016, Drug discovery today.

[44]  P. Leeson,et al.  The influence of drug-like concepts on decision-making in medicinal chemistry , 2007, Nature Reviews Drug Discovery.

[45]  Evgeny Putin,et al.  Deep biomarkers of human aging: Application of deep neural networks to biomarker development , 2016, Aging.

[46]  D. Berry.  Adaptive clinical trials in oncology , 2012, Nature Reviews Clinical Oncology.