Aim: We introduce AutoQSAR, an automated machine-learning application to build, validate and deploy quantitative structure–activity relationship (QSAR) models. Methodology/results: The process of descriptor generation, feature selection and the creation of a large number of QSAR models has been automated into a single workflow within AutoQSAR. The models are built using a variety of machine-learning methods, and each model is scored using a novel approach. Effectiveness of the method is demonstrated through comparison with literature QSAR models using identical datasets for six end points: protein–ligand binding affinity, solubility, blood–brain barrier permeability, carcinogenicity, mutagenicity and bioaccumulation in fish. Conclusion: AutoQSAR demonstrates similar or better predictive performance as compared with published results for four of the six endpoints while requiring minimal human time and expertise.

Keywords:

Papers of special note have been highlighted as: • of interest; •• of considerable interest

References

1 Li J, Ballmer SG, Gillis EP et al. Synthesis of many different types of organic small molecules using one automated process. Science 347(6227), 1221–1226 (2015).
Medline CASGoogle Scholar
2 Szymkuć S, Gajewska EP, Klucznik T et al. Computer-assisted synthetic planning: the end of the beginning. Angew. Chem. Int. Ed. Engl. 55(20), 5904–5937 (2016).
Medline CASGoogle Scholar
3 Wang L, Wu Y, Deng Y et al. Accurate and reliable prediction of relative ligand binding potency in prospective drug discovery by way of a modern free-energy calculation protocol and force field. J. Am. Chem. Soc. 137(7), 2695–2703 (2015).
Medline CASGoogle Scholar
4 Yazawa H, Hirasawa A, Horie K et al. Oxytocin receptors expressed and coupled to Ca2⁺ signalling in a human vascular smooth muscle cell line. Br. J. Pharmacol. 117(5), 799–804 (1996).
Medline CASGoogle Scholar
5 Cherkasov A, Muratov EN, Fourches D et al. QSAR modeling: where have you been? where are you going to? J. Med. Chem. 57(12), 4977–5010 (2014). • Describes the evolution of QSAR methods including development of best practices concepts and provides guidance for where the field is likely heading.
Medline CASGoogle Scholar
6 Cumming JG, Davis AM, Muresan S, Haeberlein M, Chen H. Chemical predictive modelling to improve compound quality. Nat. Rev. Drug Disc. 12(12), 948–962 (2013).
Medline CASGoogle Scholar
7 Rodgers SL, Davis AM, Tomkinson NP, van de Waterbeemd H. Predictivity of simulated ADME AutoQSAR models over time. Mol. Inform. 30(2–3), 256–266 (2011). •• Describes the experience at AstraZeneca of using QSAR models regenerated as more data became available during pharmaceutical projects.
Medline CASGoogle Scholar
8 Cox R, Green DV, Luscombe CN, Malcolm N, Pickett SD. QSAR workbench: automating QSAR modeling to drive compound design. J. Comp. Aided Mol. Des. 27(4), 321–336 (2013).
Medline CASGoogle Scholar
9 Wood DJ, Buttar D, Cumming JG, Davis AM, Norinder U, Rodgers SL. Automated QSAR with a hierarchy of global and local models. Mol. Inform. 30(11–12), 960–972 (2011). • Describes the methodology developed at AstraZeneca for automated creation of QSAR models.
Medline CASGoogle Scholar
10 Stalring JC, Carlsson LA, Almeida P, Boyer S. Azorange – high performance open source machine learning for QSAR modeling in a graphical programming environment. J. Cheminform. 3, 28 (2011).
MedlineGoogle Scholar
11 Tropsha A. Best practices for QSAR model development, validation, and exploitation. Mol. Inform. 29(6–7), 476–488 (2010). •• Outlines a best practices guidelines for QSAR model development.
Medline CASGoogle Scholar
12 Sastry M, Lowrie JF, Dixon SL, Sherman W. Large-scale systematic analysis of 2D fingerprint methods and parameters to improve virtual screening enrichments. J. Chem. Inf. Model. 50(5), 771–784 (2010).
Medline CASGoogle Scholar
13 Canvas. v2.8 Schrodinger, LLC. NY, USA (2016).
Google Scholar
14 Todeschini R, Consonni V. Handbook of Molecular Descriptors. Mannhold R, Kubinyi H, Timmerman H (Eds). Wiley-VCH, NY, USA (2008).
Google Scholar
15 Hall LH, Kier LB. Electrotopological state indices for atom types: a novel combination of electronic, topological, and valence state information. J. Chem. Inf. Comp. Sci. 35(6), 1039–1045 (1995).
CASGoogle Scholar
16 Ghose AK, Viswanadhan VN, Wendoloski JJ. Prediction of hydrophobic (lipophilic) properties of small organic molecules using fragmental methods: an analysis of Alogp and Clogp methods. J. Phys. Chem. A 102(21), 3762–3772 (1998).
CASGoogle Scholar
17 Ertl P, Rohde B, Selzer P. Fast calculation of molecular polar surface area as a sum of fragment-based contributions and its application to the prediction of drug transport properties. J. Med. Chem. 43(20), 3714–3717 (2000).
Medline CASGoogle Scholar
18 Duan J, Dixon SL, Lowrie JF, Sherman W. Analysis and comparison of 2D fingerprints: insights into database screening performance using eight fingerprint methods. J. Mol. Graph. Model. 29(2), 157–170 (2010).
Medline CASGoogle Scholar
19 An Y, Sherman W, Dixon SL. Hole filling and library optimization: application to commercially available fragment libraries. Bioorg. Med. Chem. 20(18), 5379–5387 (2012).
Medline CASGoogle Scholar
20 An Y, Sherman W, Dixon SL. Kernel-based partial least squares: application to fingerprint-based QSAR with model visualization. J. Chem. Inf. Model. 53(9), 2312–2321 (2013).
Medline CASGoogle Scholar
21 Bender A, Mussa HY, Glen RC, Reiling S. Molecular similarity searching using atom environments, information-based feature selection, and a naive bayesian classifier. J. Chem. Inf. Comp. Sci. 44(1), 170–178 (2004).
Medline CASGoogle Scholar
22 Bender A, Mussa HY, Glen RC, Reiling S. Similarity searching of chemical databases using atom environment descriptors (MOLPRINT 2D): evaluation of performance. J. Chem. Inf. Comp. Sci. 44(5), 1708–1718 (2004).
Medline CASGoogle Scholar
23 Rogers D, Brown RD, Hahn M. Using extended-connectivity fingerprints with Laplacian-modified Bayesian analysis in high-throughput screening follow-up. J. Biomol. Screen. 10(7), 682–686 (2005).
Medline CASGoogle Scholar
24 Dixon S, Merz KM Jr, Lauri G, Ianni JC. QMQSAR: utilization of a semiempirical probe potential in a field-based QSAR method. J. Comp. Chem. 26(1), 23–34 (2005).
Medline CASGoogle Scholar
25 Dixon SL, Smondyrev AM, Knoll EH, Rao SN, Shaw DE, Friesner RA. Phase: a new engine for pharmacophore perception, 3D QSAR model development, and 3D database screening: 1. Methodology and preliminary results. J. Comp. Aided Mol. Des. 20(10–11), 647–671 (2006).
Medline CASGoogle Scholar
26 Klon AE, Lowrie JF, Diller DJ. Improved naive Bayesian modeling of numerical data for absorption, distribution, metabolism and excretion (ADME) property prediction. J. Chem. Inf. Model. 46(5), 1945–1956 (2006).
Medline CASGoogle Scholar
27 Dixon SL, Villar HO. Investigation of classification methods for the prediction of activity in diverse chemical libraries. J. Comp. Aided Mol. Des. 13(5), 533–545 (1999).
Medline CASGoogle Scholar
28 Breiman L. Random forests. Machine Learning 45(1), 5–32 (2001).
Google Scholar
29 Susnow RG, Dixon SL. Use of robust classification techniques for the prediction of human cytochrome P450 2d6 Inhibition. J. Chem. Info. Comp. Sci. 43(4), 1308–1315 (2003).
Medline CASGoogle Scholar
30 Cheng A, Dixon SL. In silico models for the prediction of dose-dependent human hepatotoxicity. J. Comp.Aided Mol. Des. 17(12), 811–823 (2003).
Medline CASGoogle Scholar
31 Zhang L, Zhu H, Oprea TI, Golbraikh A, Tropsha A. QSAR modeling of the blood-brain barrier permeability for diverse organic compounds. Pharm. Res. 25(8), 1902–1914 (2008).
Medline CASGoogle Scholar
32 Wang J, Krudy G, Hou T, Zhang W, Holland G, Xu X. Development of reliable aqueous solubility models and their application in druglike analysis. J. Chem. Inf. Model. 47(4), 1395–1404 (2007).
Medline CASGoogle Scholar
33 Fjodorova N, Vracko M, Novic M, Roncaglioni A, Benfenati E. New public QSAR model for carcinogenicity. Chem. Cent. J. 4(Suppl. 1), S3 (2010).
MedlineGoogle Scholar
34 Ferrari T, Gini G. An open source multistep model to predict mutagenicity from statistical analysis and relevant structural alerts. Chem. Cent. J. 4(Suppl. 1), S2 (2010).
MedlineGoogle Scholar
35 Lombardo A, Roncaglioni A, Boriani E, Milan C, Benfenati E. Assessment and validation of the caesar predictive model for bioconcentration factor (BCF) in fish. Chem. Cent. J. 4(Suppl. 1), S1 (2010).
MedlineGoogle Scholar
36 Dimitrov S, Dimitrova N, Parkerton T, Comber M, Bonnell M, Mekenyan O. Base-line model for identifying the bioaccumulation potential of chemicals. SAR QSAR Environ. Res. 16(6), 531–554 (2005).
Medline CASGoogle Scholar
37 Zhao C, Boriani E, Chana A, Roncaglioni A, Benfenati E. A new hybrid system of QSAR models for predicting bioconcentration factors (BCF). Chemosphere 73(11), 1701–1707 (2008).
Medline CASGoogle Scholar
38 Acharjee S, Do-Rego JL, Oh DY et al. Molecular cloning, pharmacological characterization, and histochemical distribution of frog vasotocin and mesotocin receptors. J. Mol. Endocrinol. 33(1), 293–313 (2004).
Medline CASGoogle Scholar
39 Benfenati E, Benigni R, Demarini DM et al. Predictive models for carcinogenicity and mutagenicity: frameworks, state-of-the-art, and perspectives. J. Environ. Sci. Health 27(2), 57–90 (2009).
CASGoogle Scholar
40 Piegorsch WW, Zeiger E. Measuring Intra-Assay Agreement for the Ames Salmonella Assay. Hothorn L (Ed.). Springer, Berlin, Heidelberg (1991).
Google Scholar

Vol. 8, No. 15

Supplemental Materials

Metrics

Downloaded 1,076 times

History

Published online 19 September 2016

Published in print October 2016

Information

Keywords

Supplementary Material

AutoQSAR models have been provided for all end points sufficient to regenerate data from this publication. To view the supplementary material, please see the list here and the AutoQSAR models here.

Financial & competing interests disclosure

All authors are employees of Schrödinger, Inc. and some authors hold shares of that company. The authors have no other relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript apart from those disclosed.

No writing assistance was utilized in the production of this manuscript.

PDF download