DOE PAGES title logo U.S. Department of Energy
Office of Scientific and Technical Information

Title: A Machine Learning Approach for the Discovery of Ligand-Specific Functional Mechanisms of GPCRs

Abstract

G protein-coupled receptors (GPCRs) play a key role in many cellular signaling mechanisms, and must select among multiple coupling possibilities in a ligand-specific manner in order to carry out a myriad of functions in diverse cellular contexts. Much has been learned about the molecular mechanisms of ligand-GPCR complexes from Molecular Dynamics (MD) simulations. However, to explore ligand-specific differences in the response of a GPCR to diverse ligands, as is required to understand ligand bias and functional selectivity, necessitates creating very large amounts of data from the needed large-scale simulations. This becomes a Big Data problem for the high dimensionality analysis of the accumulated trajectories. Here we describe a new machine learning (ML) approach to the problem that is based on transforming the analysis of GPCR function-related, ligand-specific differences encoded in the MD simulation trajectories into a representation recognizable by state-of-the-art deep learning object recognition technology. We illustrate this method by applying it to recognize the pharmacological classification of ligands bound to the 5-HT2A and D2 subtypes of class-A GPCRs from the serotonin and dopamine families. The ML-based approach is shown to perform the classification task with high accuracy, and we identify the molecular determinants of the classifications in the contextmore » of GPCR structure and function. This study builds a framework for the efficient computational analysis of MD Big Data collected for the purpose of understanding ligand-specific GPCR activity.« less

Authors:
 [1]; ORCiD logo [1];  [2]; ORCiD logo [3]; ORCiD logo [3]
  1. Cornell Univ., Ithaca, NY (United States). Weill Medical College
  2. Cornell Univ., Ithaca, NY (United States). Weill Medical College; Consiglio Nazionale dell Ricerche, Milano (Italy). ICRM
  3. Cornell Univ., Ithaca, NY (United States). Weill Medical College; Cornell Univ., Ithaca, NY (United States). Weill Medical College. Dept. of Physiology and Biophysics. Inst. for Computational Biomedicine
Publication Date:
Research Org.:
Oak Ridge National Laboratory (ORNL), Oak Ridge, TN (United States)
Sponsoring Org.:
USDOE Office of Science (SC), Advanced Scientific Computing Research (ASCR)
OSTI Identifier:
1628488
Grant/Contract Number:  
AC05-00OR22725
Resource Type:
Accepted Manuscript
Journal Name:
Molecules
Additional Journal Information:
Journal Volume: 24; Journal Issue: 11; Journal ID: ISSN 1420-3049
Publisher:
MDPI
Country of Publication:
United States
Language:
English
Subject:
37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CHEMISTRY; 59 BASIC BIOLOGICAL SCIENCES; Biochemistry & Molecular Biology; Chemistry; Functional Selectivity; Biased Ligands; Molecular Dynamics; Deep Neural Networks; Sensitivity Analysis; Pharmacological Efficacy

Citation Formats

Plante, Ambrose, Shore, Derek M., Morra, Giulia, Khelashvili, George, and Weinstein, Harel. A Machine Learning Approach for the Discovery of Ligand-Specific Functional Mechanisms of GPCRs. United States: N. p., 2019. Web. doi:10.3390/molecules24112097.
Plante, Ambrose, Shore, Derek M., Morra, Giulia, Khelashvili, George, & Weinstein, Harel. A Machine Learning Approach for the Discovery of Ligand-Specific Functional Mechanisms of GPCRs. United States. https://doi.org/10.3390/molecules24112097
Plante, Ambrose, Shore, Derek M., Morra, Giulia, Khelashvili, George, and Weinstein, Harel. Sat . "A Machine Learning Approach for the Discovery of Ligand-Specific Functional Mechanisms of GPCRs". United States. https://doi.org/10.3390/molecules24112097. https://www.osti.gov/servlets/purl/1628488.
@article{osti_1628488,
title = {A Machine Learning Approach for the Discovery of Ligand-Specific Functional Mechanisms of GPCRs},
author = {Plante, Ambrose and Shore, Derek M. and Morra, Giulia and Khelashvili, George and Weinstein, Harel},
abstractNote = {G protein-coupled receptors (GPCRs) play a key role in many cellular signaling mechanisms, and must select among multiple coupling possibilities in a ligand-specific manner in order to carry out a myriad of functions in diverse cellular contexts. Much has been learned about the molecular mechanisms of ligand-GPCR complexes from Molecular Dynamics (MD) simulations. However, to explore ligand-specific differences in the response of a GPCR to diverse ligands, as is required to understand ligand bias and functional selectivity, necessitates creating very large amounts of data from the needed large-scale simulations. This becomes a Big Data problem for the high dimensionality analysis of the accumulated trajectories. Here we describe a new machine learning (ML) approach to the problem that is based on transforming the analysis of GPCR function-related, ligand-specific differences encoded in the MD simulation trajectories into a representation recognizable by state-of-the-art deep learning object recognition technology. We illustrate this method by applying it to recognize the pharmacological classification of ligands bound to the 5-HT2A and D2 subtypes of class-A GPCRs from the serotonin and dopamine families. The ML-based approach is shown to perform the classification task with high accuracy, and we identify the molecular determinants of the classifications in the context of GPCR structure and function. This study builds a framework for the efficient computational analysis of MD Big Data collected for the purpose of understanding ligand-specific GPCR activity.},
doi = {10.3390/molecules24112097},
journal = {Molecules},
number = 11,
volume = 24,
place = {United States},
year = {Sat Jun 01 00:00:00 EDT 2019},
month = {Sat Jun 01 00:00:00 EDT 2019}
}

Works referenced in this record:

Spontaneous Inward Opening of the Dopamine Transporter is Triggered by PIP2-Regulated Dynamics of the N-Terminus
journal, February 2016


Biased signalling: from simple switches to allosteric microprocessors
journal, January 2018

  • Smith, Jeffrey S.; Lefkowitz, Robert J.; Rajagopal, Sudarshan
  • Nature Reviews Drug Discovery, Vol. 17, Issue 4
  • DOI: 10.1038/nrd.2017.229

Crystal Structure of an LSD-Bound Human Serotonin Receptor
journal, January 2017


Scalable molecular dynamics with NAMD
journal, January 2005

  • Phillips, James C.; Braun, Rosemary; Wang, Wei
  • Journal of Computational Chemistry, Vol. 26, Issue 16, p. 1781-1802
  • DOI: 10.1002/jcc.20289

Big Data: A Survey
journal, January 2014


Spontaneous Inward Opening of the Dopamine Transporter Is Triggered by PIP 2 -Regulated Dynamics of the N-Terminus
journal, August 2015


Millisecond-scale molecular dynamics simulations on Anton
conference, January 2009

  • Shaw, David E.; Bowers, Kevin J.; Chow, Edmond
  • Proceedings of the Conference on High Performance Computing Networking, Storage and Analysis - SC '09
  • DOI: 10.1145/1654059.1654099

New insights about HERG blockade obtained from protein modeling, potential energy mapping, and docking studies
journal, May 2006

  • Farid, Ramy; Day, Tyler; Friesner, Richard A.
  • Bioorganic & Medicinal Chemistry, Vol. 14, Issue 9
  • DOI: 10.1016/j.bmc.2005.12.032

Crystal Structure of an LSD-Bound Human Serotonin Receptor
text, January 2017

  • L., Schools, Zachary; Tao, Che,; K., Shoichet, Brian
  • The University of North Carolina at Chapel Hill University Libraries
  • DOI: 10.17615/m1we-re08

Angiotensin Analogs with Divergent Bias Stabilize Distinct Receptor Conformations
journal, January 2019


High-Performance Data Analysis on the Big Trajectory Data of Cellular Scale All-atom Molecular Dynamics Simulations
journal, June 2018


Related Contribution of Specific Helix 2 and 7 Residues to Conformational Activation of the Serotonin 5-HT 2A Receptor
journal, July 1995

  • Sealfon, Stuart C.; Chi, Ling; Ebersole, Barbara J.
  • Journal of Biological Chemistry, Vol. 270, Issue 28
  • DOI: 10.1074/jbc.270.28.16683

Automation of the CHARMM General Force Field (CGenFF) I: Bond Perception and Atom Typing
journal, November 2012

  • Vanommeslaeghe, K.; MacKerell, A. D.
  • Journal of Chemical Information and Modeling, Vol. 52, Issue 12
  • DOI: 10.1021/ci300363c

CHARMM general force field: A force field for drug-like molecules compatible with the CHARMM all-atom additive biological force fields
journal, January 2009

  • Vanommeslaeghe, K.; Hatcher, E.; Acharya, C.
  • Journal of Computational Chemistry
  • DOI: 10.1002/jcc.21367

Directory of Useful Decoys, Enhanced (DUD-E): Better Ligands and Decoys for Better Benchmarking
journal, July 2012

  • Mysinger, Michael M.; Carchia, Michael; Irwin, John. J.
  • Journal of Medicinal Chemistry, Vol. 55, Issue 14
  • DOI: 10.1021/jm300687e

ZINC: A Free Tool to Discover Chemistry for Biology
journal, June 2012

  • Irwin, John J.; Sterling, Teague; Mysinger, Michael M.
  • Journal of Chemical Information and Modeling, Vol. 52, Issue 7
  • DOI: 10.1021/ci3001277

Molecular Interaction of Serotonin 5-HT 2A Receptor Residues Phe339 (6.51) and Phe340 (6.52) with Superpotent N -Benzyl Phenethylamine Agonists
journal, September 2006

  • Braden, Michael R.; Parrish, Jason C.; Naylor, John C.
  • Molecular Pharmacology, Vol. 70, Issue 6
  • DOI: 10.1124/mol.106.028720

Rapid parameterization of small molecules using the force field toolkit
journal, September 2013

  • Mayne, Christopher G.; Saam, Jan; Schulten, Klaus
  • Journal of Computational Chemistry, Vol. 34, Issue 32
  • DOI: 10.1002/jcc.23422

Automation of the CHARMM General Force Field (CGenFF) II: Assignment of Bonded Parameters and Partial Atomic Charges
journal, November 2012

  • Vanommeslaeghe, K.; Raman, E. Prabhu; MacKerell, A. D.
  • Journal of Chemical Information and Modeling, Vol. 52, Issue 12
  • DOI: 10.1021/ci3003649

Densely Connected Convolutional Networks
preprint, January 2016


Structure of the Human Dopamine D3 Receptor in Complex with a D2/D3 Selective Antagonist
journal, November 2010


Big data: Distilling meaning from data
journal, September 2008

  • Frankel, Felice; Reid, Rosalind
  • Nature, Vol. 455, Issue 7209
  • DOI: 10.1038/455030a

Protein and ligand preparation: parameters, protocols, and influence on virtual screening enrichments
journal, March 2013

  • Madhavi Sastry, G.; Adzhigirey, Matvey; Day, Tyler
  • Journal of Computer-Aided Molecular Design, Vol. 27, Issue 3
  • DOI: 10.1007/s10822-013-9644-8

Structure of the D2 dopamine receptor bound to the atypical antipsychotic drug risperidone
journal, January 2018


Structure-Function of the G Protein–Coupled Receptor Superfamily
journal, January 2013


Novel Procedure for Modeling Ligand/Receptor Induced Fit Effects
journal, January 2006

  • Sherman, Woody; Day, Tyler; Jacobson, Matthew P.
  • Journal of Medicinal Chemistry, Vol. 49, Issue 2
  • DOI: 10.1021/jm050540c

ACEMD: Accelerating Biomolecular Dynamics in the Microsecond Time Scale
journal, May 2009

  • Harvey, M. J.; Giupponi, G.; Fabritiis, G. De
  • Journal of Chemical Theory and Computation, Vol. 5, Issue 6
  • DOI: 10.1021/ct9000685

Mechanism of NMDA Receptor Channel Block by MK-801 and Memantine
journal, February 2018

  • Song, Xianqiang; Jensen, Morten Ø.; Jogini, Vishwanath
  • Biophysical Journal, Vol. 114, Issue 3
  • DOI: 10.1016/j.bpj.2017.11.180

Ligand-Dependent Conformations and Dynamics of the Serotonin 5-HT2A Receptor Determine Its Activation and Membrane-Driven Oligomerization Properties
journal, April 2012


5-HT2C Receptor Structures Reveal the Structural Basis of GPCR Polypharmacology
journal, February 2018


Densely Connected Convolutional Networks
conference, July 2017

  • Huang, Gao; Liu, Zhuang; Maaten, Laurens van der
  • 2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR)
  • DOI: 10.1109/CVPR.2017.243

CHARMM-GUI: A web-based graphical user interface for CHARMM
journal, March 2008

  • Jo, Sunhwan; Kim, Taehoon; Iyer, Vidyashankara G.
  • Journal of Computational Chemistry, Vol. 29, Issue 11
  • DOI: 10.1002/jcc.20945

Extra Precision Glide:  Docking and Scoring Incorporating a Model of Hydrophobic Enclosure for Protein−Ligand Complexes
journal, October 2006

  • Friesner, Richard A.; Murphy, Robert B.; Repasky, Matthew P.
  • Journal of Medicinal Chemistry, Vol. 49, Issue 21
  • DOI: 10.1021/jm051256o

Use of an Induced Fit Receptor Structure in Virtual Screening
journal, January 2006


Molecular Dynamics Simulation for All
journal, September 2018


The IUPHAR/BPS Guide to PHARMACOLOGY in 2018: updates and expansion to encompass the new guide to IMMUNOPHARMACOLOGY.
journalarticle, January 2018

  • Harding, Simon D.; Sharman, Joanna L.; Faccenda, Elena
  • Oxford University Press (OUP)
  • DOI: 10.17863/cam.25514

Mechanism of NMDA receptor channel block by MK-801 and memantine
journal, April 2018


Lifting the lid on GPCRs: the role of extracellular loops: GPCR extracellular loops
journal, February 2012


Revised Pharmacophore Model for 5-HT 2A Receptor Antagonists Derived from the Atypical Antipsychotic Agent Risperidone
journal, January 2019


Selectivity determinants of GPCR–G-protein binding
journal, May 2017

  • Flock, Tilman; Hauser, Alexander S.; Lund, Nadia
  • Nature, Vol. 545, Issue 7654
  • DOI: 10.1038/nature22070

Active-State Model of a Dopamine D2 Receptor - Gαi Complex Stabilized by Aripiprazole-Type Partial Agonists
journal, June 2014


A Functional Selectivity Mechanism at the Serotonin-2A GPCR Involves Ligand-Dependent Conformations of Intracellular Loop 2
journal, October 2014

  • Perez-Aguilar, Jose Manuel; Shan, Jufang; LeVine, Michael V.
  • Journal of the American Chemical Society, Vol. 136, Issue 45
  • DOI: 10.1021/ja508394x

Molecular signatures of G-protein-coupled receptors
journal, February 2013

  • Venkatakrishnan, A. J.; Deupi, Xavier; Lebon, Guillaume
  • Nature, Vol. 494, Issue 7436
  • DOI: 10.1038/nature11896

Functional Selectivity and Classical Concepts of Quantitative Pharmacology
journal, June 2006

  • Urban, Jonathan D.; Clarke, William P.; von Zastrow, Mark
  • Journal of Pharmacology and Experimental Therapeutics, Vol. 320, Issue 1
  • DOI: 10.1124/jpet.106.104463

Millisecond-scale molecular dynamics simulations on Anton
conference, January 2009

  • Shaw, David E.; Bowers, Kevin J.; Chow, Edmond
  • Proceedings of the Conference on High Performance Computing Networking, Storage and Analysis - SC '09
  • DOI: 10.1145/1654059.1654126

The ensemble nature of allostery
journal, April 2014

  • Motlagh, Hesam N.; Wrabl, James O.; Li, Jing
  • Nature, Vol. 508, Issue 7496
  • DOI: 10.1038/nature13001

Human aminolevulinate synthase structure reveals a eukaryotic-specific autoinhibitory loop regulating substrate binding and product release
journal, June 2020

  • Bailey, Henry J.; Bezerra, Gustavo A.; Marcero, Jason R.
  • Nature Communications, Vol. 11, Issue 1
  • DOI: 10.1038/s41467-020-16586-x

The IUPHAR/BPS Guide to PHARMACOLOGY in 2018: updates and expansion to encompass the new guide to IMMUNOPHARMACOLOGY
journal, November 2017

  • Harding, Simon D.; Sharman, Joanna L.; Faccenda, Elena
  • Nucleic Acids Research, Vol. 46, Issue D1
  • DOI: 10.1093/nar/gkx1121

Structural basis for Na + -sensitivity in dopamine D2 and D3 receptors
journal, January 2015

  • Michino, Mayako; Free, R. Benjamin; Doyle, Trevor B.
  • Chemical Communications, Vol. 51, Issue 41
  • DOI: 10.1039/C5CC02204E

Effector Pathway-Dependent Relative Efficacy at Serotonin Type 2A and 2C Receptors: Evidence for Agonist-Directed Trafficking of Receptor Stimulus
journal, July 1998

  • Berg, Kelly A.; Maayani, Saul; Goldfarb, Joseph
  • Molecular Pharmacology, Vol. 54, Issue 1
  • DOI: 10.1124/mol.54.1.94

A Markov State-based Quantitative Kinetic Model of Sodium Release from the Dopamine Transporter
journal, January 2017

  • Razavi, Asghar M.; Khelashvili, George; Weinstein, Harel
  • Scientific Reports, Vol. 7, Issue 1
  • DOI: 10.1038/srep40076

CHARMM-GUI: A web-based graphical user interface for CHARMM
journal, March 2008

  • Jo, Sunhwan; Kim, Taehoon; Iyer, Vidyashankara G.
  • Journal of Computational Chemistry, Vol. 29, Issue 11
  • DOI: 10.1002/jcc.20945

CHARMM general force field: A force field for drug-like molecules compatible with the CHARMM all-atom additive biological force fields
journal, January 2009

  • Vanommeslaeghe, K.; Hatcher, E.; Acharya, C.
  • Journal of Computational Chemistry
  • DOI: 10.1002/jcc.21367

Rapid parameterization of small molecules using the force field toolkit
journal, September 2013

  • Mayne, Christopher G.; Saam, Jan; Schulten, Klaus
  • Journal of Computational Chemistry, Vol. 34, Issue 32
  • DOI: 10.1002/jcc.23422

EGFRisopred: a machine learning-based classification model for identifying isoform-specific inhibitors against EGFR and HER2
journal, August 2021


Big Data: A Survey
journal, January 2014


Site-directed mutagenesis of the human dopamine D2 receptor
journal, October 1992

  • Mansour, Alfred; Meng, Fan; Meador-Woodruff, James H.
  • European Journal of Pharmacology: Molecular Pharmacology, Vol. 227, Issue 2
  • DOI: 10.1016/0922-4106(92)90129-j

New insights about HERG blockade obtained from protein modeling, potential energy mapping, and docking studies
journal, May 2006

  • Farid, Ramy; Day, Tyler; Friesner, Richard A.
  • Bioorganic & Medicinal Chemistry, Vol. 14, Issue 9
  • DOI: 10.1016/j.bmc.2005.12.032

Spontaneous Inward Opening of the Dopamine Transporter is Triggered by PIP2-Regulated Dynamics of the N-Terminus
journal, February 2016


Mechanism of NMDA Receptor Channel Block by MK-801 and Memantine
journal, February 2018

  • Song, Xianqiang; Jensen, Morten Ø.; Jogini, Vishwanath
  • Biophysical Journal, Vol. 114, Issue 3
  • DOI: 10.1016/j.bpj.2017.11.180

Angiotensin Analogs with Divergent Bias Stabilize Distinct Receptor Conformations
journal, January 2019


Molecular Dynamics Simulation for All
journal, September 2018


Revised Pharmacophore Model for 5-HT 2A Receptor Antagonists Derived from the Atypical Antipsychotic Agent Risperidone
journal, January 2019


ZINC: A Free Tool to Discover Chemistry for Biology
journal, June 2012

  • Irwin, John J.; Sterling, Teague; Mysinger, Michael M.
  • Journal of Chemical Information and Modeling, Vol. 52, Issue 7
  • DOI: 10.1021/ci3001277

Automation of the CHARMM General Force Field (CGenFF) II: Assignment of Bonded Parameters and Partial Atomic Charges
journal, November 2012

  • Vanommeslaeghe, K.; Raman, E. Prabhu; MacKerell, A. D.
  • Journal of Chemical Information and Modeling, Vol. 52, Issue 12
  • DOI: 10.1021/ci3003649

Novel Procedure for Modeling Ligand/Receptor Induced Fit Effects
journal, January 2006

  • Sherman, Woody; Day, Tyler; Jacobson, Matthew P.
  • Journal of Medicinal Chemistry, Vol. 49, Issue 2
  • DOI: 10.1021/jm050540c

Extra Precision Glide:  Docking and Scoring Incorporating a Model of Hydrophobic Enclosure for Protein−Ligand Complexes
journal, October 2006

  • Friesner, Richard A.; Murphy, Robert B.; Repasky, Matthew P.
  • Journal of Medicinal Chemistry, Vol. 49, Issue 21
  • DOI: 10.1021/jm051256o

Directory of Useful Decoys, Enhanced (DUD-E): Better Ligands and Decoys for Better Benchmarking
journal, July 2012

  • Mysinger, Michael M.; Carchia, Michael; Irwin, John. J.
  • Journal of Medicinal Chemistry, Vol. 55, Issue 14
  • DOI: 10.1021/jm300687e

Big data: Distilling meaning from data
journal, September 2008

  • Frankel, Felice; Reid, Rosalind
  • Nature, Vol. 455, Issue 7209
  • DOI: 10.1038/455030a

Molecular signatures of G-protein-coupled receptors
journal, February 2013

  • Venkatakrishnan, A. J.; Deupi, Xavier; Lebon, Guillaume
  • Nature, Vol. 494, Issue 7436
  • DOI: 10.1038/nature11896

The ensemble nature of allostery
journal, April 2014

  • Motlagh, Hesam N.; Wrabl, James O.; Li, Jing
  • Nature, Vol. 508, Issue 7496
  • DOI: 10.1038/nature13001

Selectivity determinants of GPCR–G-protein binding
journal, May 2017

  • Flock, Tilman; Hauser, Alexander S.; Lund, Nadia
  • Nature, Vol. 545, Issue 7654
  • DOI: 10.1038/nature22070

Biased signalling: from simple switches to allosteric microprocessors
journal, January 2018

  • Smith, Jeffrey S.; Lefkowitz, Robert J.; Rajagopal, Sudarshan
  • Nature Reviews Drug Discovery, Vol. 17, Issue 4
  • DOI: 10.1038/nrd.2017.229

A novel specific PERK activator reduces toxicity and extends survival in Huntington's disease models
journal, April 2020


A Markov State-based Quantitative Kinetic Model of Sodium Release from the Dopamine Transporter
journal, January 2017

  • Razavi, Asghar M.; Khelashvili, George; Weinstein, Harel
  • Scientific Reports, Vol. 7, Issue 1
  • DOI: 10.1038/srep40076

Related Contribution of Specific Helix 2 and 7 Residues to Conformational Activation of the Serotonin 5-HT 2A Receptor
journal, July 1995

  • Sealfon, Stuart C.; Chi, Ling; Ebersole, Barbara J.
  • Journal of Biological Chemistry, Vol. 270, Issue 28
  • DOI: 10.1074/jbc.270.28.16683

High-Performance Data Analysis on the Big Trajectory Data of Cellular Scale All-atom Molecular Dynamics Simulations
journal, June 2018


Lifting the lid on GPCRs: the role of extracellular loops: GPCR extracellular loops
journal, February 2012


Use of an Induced Fit Receptor Structure in Virtual Screening
journal, January 2006


Functional Selectivity and Classical Concepts of Quantitative Pharmacology
journal, June 2006

  • Urban, Jonathan D.; Clarke, William P.; von Zastrow, Mark
  • Journal of Pharmacology and Experimental Therapeutics, Vol. 320, Issue 1
  • DOI: 10.1124/jpet.106.104463

Structure of the Human Dopamine D3 Receptor in Complex with a D2/D3 Selective Antagonist
journal, November 2010


Millisecond-scale molecular dynamics simulations on Anton
conference, January 2009

  • Shaw, David E.; Bowers, Kevin J.; Chow, Edmond
  • Proceedings of the Conference on High Performance Computing Networking, Storage and Analysis - SC '09
  • DOI: 10.1145/1654059.1654126

Structure-Function of the G Protein–Coupled Receptor Superfamily
journal, January 2013


Crystal Structure of an LSD-Bound Human Serotonin Receptor
text, January 2017

  • L., Schools, Zachary; Tao, Che,; K., Shoichet, Brian
  • The University of North Carolina at Chapel Hill University Libraries
  • DOI: 10.17615/m1we-re08

DSNet for Real-Time Driving Scene Semantic Segmentation
preprint, January 2018