A renaissance of neural networks in drug discovery

Expert Opinion on Drug Discovery

Volumn 11, Issue 8, 2016, Pages 785-795

  Baskin, Igor I ^a,b   Winkler, David ^c,d,e,f   Tetko, Igor V ^g,h  

^a MOSCOW STATE UNIVERSITY   (Russian Federation)

^b KAZAN FEDERAL UNIVERSITY   (Russian Federation)

^c CSIRO   (Australia)

^d MONASH UNIVERSITY   (Australia)

^e LA TROBE UNIVERSITY   (Australia)

^f FLINDERS UNIVERSITY   (Australia)

^g INSTITUTE OF EPIDEMIOLOGY   (Germany)

^h BigChem GmbH   (Germany)

Author keywords

Deep learning; neural network ensembles; neural networks; overfitting; structure activity relationships

Indexed keywords

1,4 DIHYDROPYRIDINE DERIVATIVE; CALCIUM CHANNEL BLOCKING AGENT; CLOTRIMAZOLE; ENDOCRINE DISRUPTOR; INHALATION ANESTHETIC AGENT; NONNUCLEOSIDE REVERSE TRANSCRIPTASE INHIBITOR; PROTEIN FARNESYLTRANSFERASE INHIBITOR; DRUG;

ARTIFICIAL NEURAL NETWORK; BACK PROPAGATION; BAYESIAN LEARNING; CONSENSUS; DRUG ABSORPTION; DRUG DELIVERY SYSTEM; DRUG DESIGN; DRUG DISTRIBUTION; DRUG EXCRETION; DRUG METABOLISM; DRUG SAFETY; DRUG SCREENING; DRUG SELECTIVITY; HUMAN; IC50; MACHINE LEARNING; MOLECULAR BIOLOGY; PERCEPTRON; PHYSICAL CHEMISTRY; PRIORITY JOURNAL; PRODUCT DEVELOPMENT; QUANTITATIVE STRUCTURE ACTIVITY RELATION; QUANTUM CHEMISTRY; REVIEW; SELF ORGANIZING MAP; STATISTICAL MODEL; VIRTUAL REALITY; BIOLOGICAL MODEL; CHEMISTRY; COMPUTER AIDED DESIGN; DRUG DEVELOPMENT; PROCEDURES; STRUCTURE ACTIVITY RELATION;

COMPUTER-AIDED DESIGN; DRUG DELIVERY SYSTEMS; DRUG DESIGN; DRUG DISCOVERY; HUMANS; MODELS, BIOLOGICAL; NEURAL NETWORKS (COMPUTER); PHARMACEUTICAL PREPARATIONS; STRUCTURE-ACTIVITY RELATIONSHIP;

EID: 84978792062     PISSN: 17460441     EISSN: 1746045X     Source Type: Journal    
DOI: 10.1080/17460441.2016.1201262     Document Type: Review

References (110)

1. W.S.McCulloch, W.Pitts A logical calculus of the ideas immanent in nervous activity. Bull Math Biophys. 1943;5:115–133. doi:10.1007/BF02478259.
2. F.Rosenblatt. The perceptron: a probabilistic model for information storage and organization in the brain. Psychol Rev. 1958;65(6):386–408.
3. D.E.Rumelhart, G.E.Hinton, R.J.Williams. Learning internal representations by back-propagating errors. Nature. 1986;33:533–536. doi:10.1038/323533a0.•• The main reference for backpropagation neural networks.
4. K.Fukushima. Neocognitron: a self-organizing neural network model for a mechanism of pattern recognition unaffected by shift in position. Biol Cybernetics. 1980;36:193–202. doi:10.1007/BF00344251.
5. T.Kohonen. Self-organizing maps. Berlin: Springer; 2001.•• The main reference on self-organizing maps.
6. J.J.Hopfield. Neural networks and physical systems with emergent collective computational abilities. Proc Natl Acad Sci U S A. 1982;79(8):2554–2558.
7. G.E.Hinton, R.R.Salakhutdinov. Reducing the dimensionality of data with neural networks. Science. 2006;313(5786):504–507. doi:10.1126/science.1127647.•• First publication on deep learning.
8. Y.Bengio. Learning deep architectures for AI. Foundations Trends Machine Learning. 2009;2(1):1–127. doi:10.1561/2200000006.
9. Y.LeCun, Y.Bengio, G.Hinton. Deep learning. Nature. 2015;521(7553):436–444. doi:10.1038/nature14539.•• Important recent review on deep learning.
10. S.A.Hiller, V.E.Golender, A.B.Rosenblit, et al. Cybernetic methods of drug design. I. Statement of the problem – the perceptron approach. Comput Biomed Res. 1973;6(5):411–421.• First publication on the use of neural networks in drug discovery.
11. T.Aoyama, Y.Suzuki, H.Ichikawa. Neural networks applied to structure-activity relationships. J Med Chem. 1990;33(3):905–908.
12. D.A.Winkler, F.R.Burden. Application of neural networks to large dataset QSAR, virtual screening, and library design. Methods Mol Biol. 2002;201:325–367. doi:10.1385/1-59259-285-6:325.
13. N.M.Halberstam, I.I.Baskin, V.A.Palyulin, et al. Neural networks as a method for elucidating structure-property relationships for organic compounds. Russian Chem Rev. 2003;72(7):629–649. doi:10.1070/RC2003v072n07ABEH000754.
14. I.I.Baskin, V.A.Palyulin, N.S.Zefirov. Neural networks in building QSAR models. Methods Mol Biol (Clifton, NJ). 2008;458:137–158.• Review concerning the use of neural network in QSAR modeling.
15. V.G.Maltarollo, D.S.Abf, K.M.Honório. Applications of artificial neural networks in chemical problems. In: K.Suzuki, editor. Artificial neural networks-architectures and applications. INTECH; 2013. p. 203–223. doi:10.5772/3409.
16. J.C.Dearden, P.H.Rowe. Use of artificial neural networks in the QSAR prediction of physicochemical properties and toxicities for REACH legislation. Methods Mol Biol (Clifton, NJ). 2015;1260:65–88.
17. I.I.Baskin, M.I.Skvortsova, I.V.Stankevich, et al. On the basis of invariants of labeled molecular graphs. J Chem Inf Comput Sci. 1995;35(3):527–531.
18. N.V.Artemenko, I.I.Baskin, V.A.Palyulin, et al. Artificial neural network and fragmental approach in prediction of physicochemical properties of organic compounds. Russian Chem Bull. 2003;52(1):20–29. doi:10.1023/A:1022467508832.
19. D.E.Rumelhart, G.E.Hinton, R.J.Williams. Learning internal representation by error propagation. In: D.E.Rumelhart, J.L.McClelland, editors. Parallel distributed processing: explorations in the microstructure of cognition, volume 1: foundations. Cambridge (MA): MIT Press; 1986. p. 318–362.
20. M.Riedmiller, H.Braun. A direct adaptive method for faster backpropagation learning: the RPROP algorithm. Proc IEEE Int Conf Neural Networks. 1993;586–591.
21. M.T.Hagan, M.Menhaj. Training feedforward networks with the Marquardt algorithm. IEEE Trans Neural Networks. 1994;5(6):989–993. doi:10.1109/72.329697.
22. I.V.Tetko, D.J.Livingstone, A.I.Luik. Neural network studies. 1. Comparison of overfitting and overtraining. J Chem Inf Comput Sci. 1995;35(5):826–833.•• Method to avoid overfitting of neural networks, notwithstanding of their complexity, by early stopping.
23. N.Srebro, A.Shraibman. Rank, trace-norm and max-norm. In: Learning theory. Berlin: Springer; 2005. p. 545–560.
24. F.Burden, D.Winkler. Bayesian regularization of neural networks. Methods Mol Biol. 2008;458:25–44.
25. F.R.Burden, D.A.Winkler. Robust QSAR models using Bayesian regularized neural networks. J Med Chem. 1999;42(16):3183–3187. doi:10.1021/jm980697n.•• First application of Bayesian regularized neural networks in QSAR modeling.
26. F.R.Burden, D.A.Winkler. An optimal self-pruning neural network and nonlinear descriptor selection in QSAR. QSAR Comb Sci. 2009;28(10):1092–1097. doi:10.1002/qsar.v28:10.
27. N.Srivastava, G.Hinton, A.Krizhevsky, et al. Dropout: a simple way to prevent neural networks from overfitting. J Machine Learn Res. 2014;15(1):1929–1958.• Important method to avoid overfitting used in deep-learning neural network.
28. A.Varnek, C.Gaudin, G.Marcou, et al. Inductive transfer of knowledge: application of multi-task learning and feature net approaches to model tissue-air partition coefficients. J Chem Inf Model. 2009;49(1):133–144. doi:10.1021/ci8002914.
29. R.Caruana. Multitask learning. Mach Learn. 1997;28(1):41–75. doi:10.1023/A:1007379606734.
30. I.I.Baskin, N.M.Halberstam, T.V.Mukhina, et al. The learned symmetry concept in revealing quantitative structure-activity relationships with artificial neural networks. SAR QSAR Environ Res. 2001;12(4):401–416. doi:10.1080/10629360108033247.
31. I.V.Tetko, A.E.Villa, D.J.Livingstone. Neural network studies. 2. Variable selection. J Chem Inf Comput Sci. 1996;36(4):794–803.
32. I.I.Baskin, A.O.Ait, N.M.Halberstam, et al. An approach to the interpretation of backpropagation neural network models in QSAR studies. SAR QSAR Environ Res. 2002;13(1):35–41. doi:10.1080/10629360290002073.
33. M.Markou, S.Singh. Novelty detection: a review – part 2: neural network based approaches. Signal Process. 2003;83(12):2499–2521. doi:10.1016/j.sigpro.2003.07.019.
34. P.V.Karpov, D.I.Osolodkin, I.I.Baskin, et al. One-class classification as a novel method of ligand-based virtual screening: the case of glycogen synthase kinase 3ОІ inhibitors. Bioorg Med Chem Lett. 2011;21(22):6728–6731. doi:10.1016/j.bmcl.2011.09.051.
35. A.N.Tikhonov, V.Y.Arsenin. Solutions of ill-posed problems. New York: Winston; 1977.
36. F.R.Burden, M.G.Ford, D.C.Whitley, et al. Use of automatic relevance determination in QSAR studies using Bayesian neural networks. J Chem Inf Comput Sci. 2000;40(6):1423–1430.
37. F.R.Burden, D.A.Winkler. Optimal sparse descriptor selection for QSAR using Bayesian methods. QSAR Comb Sci. 2009;28(6–7):645–653. doi:10.1002/qsar.v28:6/7.• First description of the use of feature selection using sparse Bayesian priors.
38. D.A.Winkler, F.R.Burden. Bayesian neural nets for modeling in drug discovery. Drug Discovery Today BIOSILICO. 2004;2(3):104–111. doi:10.1016/S1741-8364(04)02393-5.
39. D.J.C.Mackay. A practical Bayesian framework for backpropagation networks. Neural Comput. 1992;4(3):448–472. doi:10.1162/neco.1992.4.3.448.
40. C.M.Bishop. Neural networks for pattern recognition. Oxford (UK): Oxford University Press; 1995.
41. M.A.T.Figueiredo. Adaptive sparseness for supervised learning. IEEE Trans Pattern Anal. 2003;25(9):1150–1159. doi:10.1109/TPAMI.2003.1227989.• Describes the mathematics involved in Bayesian regularization for regression.
42. P.Bruneau. Search for predictive generic model of aqueous solubility using Bayesian neural nets. J Chem Inf Comput Sci. 2001;41(6):1605–1616.
43. P.Bruneau, N.R.McElroy. logD(7.4) modeling using Bayesian regularized neural networks. Assessment and correction of the errors of prediction. J Chem Inf Model. 2006;46(3):1379–1387. doi:10.1021/ci0504014.
44. F.R.Burden, D.A.Winkler. A quantitative structure-activity relationships model for the acute toxicity of substituted benzenes to Tetrahymena pyriformis using Bayesian-regularized neural networks. Chem Res Toxicol. 2000;13(6):436–440.
45. F.R.Burden, D.A.Winkler. Predictive Bayesian neural network models of MHC class II peptide binding. J Mol Graph Model. 2005;23(6):481–489. doi:10.1016/j.jmgm.2005.03.001.
46. D.T.Manallack, F.R.Burden, D.A.Winkler. Modelling inhalational anaesthetics using Bayesian feature selection and QSAR modelling methods. ChemMedChem. 2010;5(8):1318–1323. doi:10.1002/cmdc.201000056.
47. J.Ma, R.P.Sheridan, A.Liaw, et al. Deep neural nets as a method for quantitative structure-activity relationships. J Chem Inf Model. 2015;55(2):263–274. doi:10.1021/ci500747n.
48. M.Salahinejad, T.C.Le, D.A.Winkler. Aqueous solubility prediction: do crystal lattice interactions help? Mol Pharmaceut. 2013;10(7):2757–2766. doi:10.1021/mp4001958.
49. R.Orre, A.Bate, M.Lindquist. Bayesian neural networks used to find adverse drug combinations and drug related syndromes. Persp Neural Comp. 2000;215–220. doi:10.1007/978-1-4471-0513-8_32.
50. D.A.Winkler, F.R.Burden. Modelling blood-brain barrier partitioning using Bayesian neural nets. J Mol Graph Model. 2004;22(6):499–505. doi:10.1016/j.jmgm.2004.03.010.
51. M.J.Polley, F.R.Burden, D.A.Winkler. Predictive human intestinal absorption QSAR models using Bayesian regularized neural networks. Aust J Chem. 2005;58(12):859–863. doi:10.1071/CH05202.
52. M.J.Polley, D.A.Winkler, F.R.Burden. Broad-based quantitative structure-activity relationship modeling of potency and selectivity of farnesyltransferase inhibitors using a Bayesian regularized neural network. J Med Chem. 2004;47(25):6230–6238. doi:10.1021/jm049621j.
53. 2+-activated K+ channel by some triarylmethanes using topological charge indexes descriptors. J Comput Aid Mol Des. 2005;19(11):771–789. doi:10.1007/s10822-005-9025-z.
54. M.Fernandez, J.Caballero, L.Fernandez, et al. Genetic algorithm optimization in drug design QSAR: Bayesian-regularized genetic neural networks (BRGNN) and genetic algorithm-optimized support vectors machines (GA-SVM). Mol Divers. 2011;15(1):269–289. doi:10.1007/s11030-010-9234-9.
55. V.C.Epa, F.R.Burden, C.Tassa, et al. Modeling biological activities of nanoparticles. Nano Lett. 2012;12(11):5808–5812. doi:10.1021/nl303144k.
56. D.A.Winkler, E.Mombelli, A.Pietroiusti, et al. Applying quantitative structure-activity relationship approaches to nanotoxicology: current status and future potential. Toxicol. 2013;313(1):15–23. doi:10.1016/j.tox.2012.11.005.
57. T.C.Le, B.Yan, D.A.Winkler. Robust prediction of personalized cell recognition from a cancer population by a dual targeting nanoparticle library. Adv Funct Mater. 2015;25(44):6927–6935. doi:10.1002/adfm.201502811.
58. I.V.Tetko, A.I.Luik, G.I.Poda. Applications of neural networks in structure-activity relationships of a small number of molecules. J Med Chem. 1993;36(7):811–814.• First use of neural network ensembles in drug discovery.
59. L.Breiman. Random forests. Machine Learn. 2001;45(1):5–32. doi:10.1023/A:1010933404324.
60. J.R.Votano, M.Parham, L.H.Hall, et al. Three new consensus QSAR models for the prediction of Ames genotoxicity. Mutagenesis. 2004;19(5):365–377. doi:10.1093/mutage/geh043.
61. J.H.Tomal, W.J.Welch, R.H.Zamar. Exploiting multiple descriptor sets in QSAR studies. J Chem Inf Model. 2016;56(3):501–509. doi:10.1021/acs.jcim.5b00663.
62. S.Novotarskyi, I.Sushko, R.Korner, et al. A comparison of different QSAR approaches to modeling CYP450 1A2 inhibition. J Chem Inf Model. 2011;51(6):1271–1280. doi:10.1021/ci200091h.
63. B.Nizami, I.V.Tetko, N.A.Koorbanally, et al. QSAR models and scaffold-based analysis of non-nucleoside HIV RT inhibitors. Chemometrics Intell Lab Syst. 2015;148:134–144. doi:10.1016/j.chemolab.2015.09.011.
64. K.Mansouri, A.Abdelaziz, A.Rybacka, et al. CERAPP: collaborative estrogen receptor activity prediction project. Environ Health Perspect. 2016. doi:10.1289/ehp.1510267.
65. I.V.Tetko, I.Sushko, A.K.Pandey, et al. Critical assessment of QSAR models of environmental toxicity against Tetrahymena pyriformis: focusing on applicability domain and overfitting by variable selection. J Chem Inf Model. 2008;48(9):1733–1746. doi:10.1021/ci800151m.
66. I.Sushko, S.Novotarskyi, R.Körner, et al. Applicability domain for in silico models to achieve accuracy of experimental measurements. J Chemom. 2010;24(3–4):202–208. doi:10.1002/cem.1296.
67. A.E.Villa, I.V.Tetko, P.Dutoit, et al. Corticofugal modulation of functional connectivity within the auditory thalamus of rat, guinea pig and cat revealed by cooling deactivation. J Neurosci Methods. 1999;86(2):161–178.
68. I.V.Tetko, V.Y.Tanchuk. Application of associative neural networks for prediction of lipophilicity in ALOGPS 2.1 program. J Chem Inf Comput Sci. 2002;42(5):1136–1145.
69. I.V.Tetko, I.Jaroszewicz, J.A.Platts, et al. Calculation of lipophilicity for Pt(II) complexes: experimental comparison of several methods. J Inorg Biochem. 2008;102(7):1424–1437. doi:10.1016/j.jinorgbio.2007.12.029.
70. I.V.Tetko, G.I.Poda. Application of ALOGPS 2.1 to predict log D distribution coefficient for Pfizer proprietary compounds. J Med Chem. 2004;47(23):5601–5604. doi:10.1021/jm049509l.
71. I.V.Tetko, G.I.Poda, C.Ostermann, et al. Large-scale evaluation of log P predictors: local corrections may compensate insufficient accuracy and need of experimentally testing every other compound. Chem Biodivers. 2009;6(11):1837–1844. doi:10.1002/cbdv.200900075.
72. I.V.Tetko, Y.Sushko, S.Novotarskyi, et al. How accurately can we predict the melting points of drug-like compounds? J Chem Inf Model. 2014;54(12):3320–3329. doi:10.1021/ci5005288.
73. I.V.Tetko, H.P.Varbanov, M.Galanski, et al. Prediction of logP for Pt(II) and Pt(IV) complexes: Comparison of statistical and quantum-chemistry based approaches. J Inorg Biochem. 2016;156:1–13. doi:10.1016/j.jinorgbio.2015.12.006.
74. S.Novotarskyi, A.Abdelaziz, Y.Sushko, et al. ToxCast EPA in vitro to in vivo challenge: insight into the Rank-i model. Chem Res Toxicol. 2016;29(5):768–775. doi:10.1021/acs.chemrestox.5b00481.
75. A.Abdelaziz, H.Spahn-Langguth, K.Werner-Schramm, et al. Consensus modeling for HTS assays using in silico descriptors calculates the best balanced accuracy in Tox21 challenge. Frontiers Environ Sci. 2016;4(2). doi:10.3389/fenvs.2016.00002.
76. S.Anzali, J.Gasteiger, U.Holzgrabe, et al. The use of self-organizing neural networks in drug design. Perspect Drug Discov Des. 1998;9:273–299. doi:10.1023/A:1027276425268.
77. J.Polanski, R.Gieleciak, A.Bak. The comparative molecular surface analysis (COMSA) – a nongrid 3D QSAR method by a coupled neural network and PLS system: predicting pKa values of benzoic and alkanoic acids. J Chem Inf Comput Sci. 2002;42(2):184–191.
78. I.V.Tetko, V.V.Kovalishyn, D.J.Livingstone. Volume learning algorithm artificial neural networks for 3D QSAR studies. J Med Chem. 2001;44(15):2411–2420.
79. M.Wang, L.Li, C.Yu, et al. Classification of mixtures of Chinese herbal medicines based on a Self-Organizing Map (SOM). Mol Inform. 2016. doi:10.1002/minf.201500115.
80. P.Schneider, A.T.Mueller, G.Gabernet, et al. Hybrid network model for “deep learning” of chemical data: application to antimicrobial peptides. Mol Inform. 2016. doi:10.1002/minf.201600011.
81. H.A.Gaspar, I.I.Baskin, G.Marcou, et al. Chemical data visualization and analysis with incremental generative topographic mapping: big data challenge. J Chem Inf Model. 2015;55(1):84–94. doi:10.1021/ci500575y.
82. H.A.Gaspar, I.I.Baskin, G.Marcou, et al. GTM-based QSAR models and their applicability domains. Mol Inform. 2015;34(6–7):348–356. doi:10.1002/minf.201400153.
83. H.A.Gaspar, G.Marcou, D.Horvath, et al. Generative topographic mapping-based classification models and their applicability domain: application to the Biopharmaceutics Drug Disposition Classification System (BDDCS). J Chem Inf Model. 2013;53(12):3318–3325. doi:10.1021/ci400423c.
84. H.A.Gaspar, I.I.Baskin, G.Marcou, et al. Stargate GTM: bridging descriptor and activity spaces. J Chem Inf Model. 2015;55(11):2403–2410. doi:10.1021/acs.jcim.5b00398.
85. I.I.Baskin, V.A.Palyulin, N.S.Zefirov. A neural device for searching direct correlations between structures and properties of chemical compounds. J Chem Inf Comput Sci. 1997;37(4):715–721. doi:10.1021/ci940128y.• First neural network with discrete convolutional architecture for searching direct correlations between structures and properties of chemical compounds.
86. A.Micheli, A.Sperduti, A.Starita, et al. Analysis of the internal representations developed by neural networks for structures applied to quantitative structure-activity relationship studies of benzodiazepines. J Chem Inf Comput Sci. 2001;41(1):202–218.
87. A.Goulon, T.Picot, A.Duprat, et al. Predicting activities without computing descriptors: graph machines for QSAR. SAR QSAR Environ Res. 2007;18(1–2):141–153. doi:10.1080/10629360601054313.
88. T.C.Le, C.E.Conn, B.Fr, et al. Computational modeling and prediction of the complex time-dependent phase behavior of lyotropic liquid crystals under in meso crystallization conditions. Cryst Growth Des. 2013;13(3):1267–1276. doi:10.1021/cg301730z.
89. R.Orre, A.Bate, G.N.Noren, et al. A Bayesian recurrent neural network for unsupervised pattern recognition in large incomplete data sets. Int J Neural Syst. 2005;15(3):207–222. doi:10.1142/S0129065705000219.
90. W.Y.Goh, C.P.Lim, K.K.Peh, et al. Application of a recurrent neural network to prediction of drug dissolution profiles. Neural Comput Appl. 2002;10(4):311–317. doi:10.1007/s005210200003.
91. I.Bonet, M.M.Garcia, Y.Saeys, et al. Predicting human immunodeficiency virus (HIV) drug resistance using recurrent neural networks. In: J.Mira, J.R.Alvarez, editors. Bio-inspired modeling of cognitive tasks, pt 1, proceedings. 2007. Berlin: Springer. p. 234–243.
92. Y.Bengio, A.Courville, P.Vincent. Representation learning: a review and new perspectives. IEEE Trans Pattern Anal Mach Intell. 2013;35(8):1798–1828. doi:10.1109/TPAMI.2013.50.• Important review on representation learning.
93. X.Glorot, A.Bordes, Y.Bengio. Deep sparse rectifier neural networks. Int Conf Artif Intelligence Stat. 2011;2011:315–323.
94. Y.LeCun, L.Bottou, Y.Bengio, et al. Gradient-based learning applied to document recognition. Proc IEEE. 1998;86(11):2278–2324. doi:10.1109/5.726791.
95. S.J.Pan, Q.Yang. A survey on transfer learning. IEEE Trans Knowl Data Eng. 2010;22(10):1345–1359. doi:10.1109/TKDE.2009.191.
96. O.Chapelle, B.Schoelkopf, A.Zien. Semi-supervised learning. Cambridge (MA): The MIT Press; 2006.
97. A.Varnek, I.I.Baskin. Chemoinformatics as a theoretical chemistry discipline. Mol Inform. 2011;30(1):20–32. doi:10.1002/minf.v30.1.• An important review on the use of machine learning in structure–activity modeling
98. A.Varnek, I.Baskin. Machine learning methods for property prediction in chemoinformatics: quo vadis? J Chem Inf Model. 2012;52(6):1413–1437. doi:10.1021/ci200409x.
99. A.Cherkasov, E.N.Muratov, D.Fourches, et al. QSAR modeling: where have you been? Where are you going to? J Med Chem. 2015;57(12):4977–5010. doi:10.1021/jm4004285.•• Comprehensive review on QSAR modeling
100. A.Lusci, G.Pollastri, P.Baldi Deep architectures and deep learning in chemoinformatics: the prediction of aqueous solubility for drug-like molecules. J Chem Inf Model. 2013;53(7):1563–1575.
101. B.Ramsundar, S.Kearnes, P.Riley, et al. Massively multitask networks for drug discovery. arXiv preprint arXiv:150202072. 2015. [cited 2016 Jun 10]. Available from: https://arxiv.org/abs/1502.02072
102. E.Gawehn, J.A.Hiss, G.Schneider. Deep learning in drug discovery. Mol Inf. 2016;35(1):3–14. doi:10.1002/minf.v35.1.
103. J.Markoff Scientists see promise in deep-learning programs. The New York Times. 2012 Nov 23.
104. G.E.Dahl, N.Jaitly, R.Salakhutdinov. Multi-task neural networks for QSAR predictions. arXiv preprint arXiv:14061231. 2014. [cited 2016 Jun 10]. Available from: http://arxiv.org/abs/1406.1231
105. T.Unterthiner, A.Mayr, G.Ünter Klambauer, et al. Deep learning as an opportunity in virtual screening. In: Deep Learning and Representation Learning Workshop, NIPS. 2014.
106. T.Unterthiner, A.Mayr, G.Klambauer, et al. Toxicity prediction using deep learning. arXiv preprint arXiv:150301445. 2015. [cited 2016 Jun 10]. Available from: https://arxiv.org/abs/1503.01445
107. Y.Sushko, S.Novotarskyi, R.Korner, et al. Prediction-driven matched molecular pairs to interpret QSARs and aid the molecular optimization process. J Cheminform. 2014;6(1):48. doi:10.1186/1758-2946-6-6.
108. I.V.Tetko, O.Engkvist, U.Koch, et al. BIGCHEM: challenges and opportunities for Big Data analysis in chemistry. Mol Inf. 2016.•• An important review on Big Data in chemistry.
109. M.Cruz-Monteagudo, J.Medina-Franco, Y.Pérez-Castillo, et al. Activity cliffs in drug discovery: Dr Jekyll or Mr Hyde? Drug Discov Today. 2014;19(8):1069–1080. doi:10.1016/j.drudis.2014.02.003.
110. N.V.Kireeva, S.I.Ovchinnikova, S.L.Kuznetsov, et al. Impact of distance-based metric learning on classification and visualization model performance and structure-activity landscapes. J Comput Aid Mol Des. 2014;28(2):61–73. doi:10.1007/s10822-014-9719-1.