Efficient nonmyopic active search with applications in drug and materials discovery

Active search is a learning paradigm for actively identifying as many members of a given class as possible. A critical target scenario is high-throughput screening for scientific discovery, such as drug or materials discovery. In this paper, we approach this problem in Bayesian decision framework. We first derive the Bayesian optimal policy under a natural utility, and establish a theoretical hardness of active search, proving that the optimal policy can not be approximated for any constant ratio. We also study the batch setting for the first time, where a batch of $b>1$ points can be queried at each iteration. We give an asymptotic lower bound, linear in batch size, on the adaptivity gap: how much we could lose if we query $b$ points at a time for $t$ iterations, instead of one point at a time for $bt$ iterations. We then introduce a novel approach to nonmyopic approximations of the optimal policy that admits efficient computation. Our proposed policy can automatically trade off exploration and exploitation, without relying on any tuning parameters. We also generalize our policy to batch setting, and propose two approaches to tackle the combinatorial search challenge. We evaluate our proposed policies on a large database of drug discovery and materials science. Results demonstrate the superior performance of our proposed policy in both sequential and batch setting; the nonmyopic behavior is also illustrated in various aspects.

[1]  Barnabás Póczos,et al.  Active learning and search on low-rank matrices , 2013, KDD.

[2]  Andreas Krause,et al.  Near-optimal Batch Mode Active Learning and Adaptive Submodular Optimization , 2013, ICML.

[3]  Harold J. Kushner,et al.  A New Method of Locating the Maximum Point of an Arbitrary Multipeak Curve in the Presence of Noise , 1964 .

[4]  Roman Garnett,et al.  Efficient nonmyopic batch active search , 2018, NeurIPS.

[5]  Roman Garnett,et al.  Introducing the ‘active search’ method for iterative virtual screening , 2015, Journal of Computer-Aided Molecular Design.

[6]  Jialei Wang,et al.  Bayesian Optimization with Parallel Function Evaluations and Multiple Information Sources: Methodology with Applications in Biochemistry, Aerospace Engineering, and Machine Learning , 2017 .

[7]  Roman Garnett,et al.  Efficient Nonmyopic Active Search , 2017, ICML.

[8]  Neil D. Lawrence,et al.  Batch Bayesian Optimization via Local Penalization , 2015, AISTATS.

[9]  Roman Garnett,et al.  Active Search for Computer‐aided Drug Design , 2018, Molecular informatics.

[10]  Alok Choudhary,et al.  A General-Purpose Machine Learning Framework for Predicting Properties of Inorganic Materials , 2016 .

[11]  M. L. Fisher,et al.  An analysis of approximations for maximizing submodular set functions—I , 1978, Math. Program..

[12]  Roman Garnett,et al.  Active Search in Intensionally Specified Structured Spaces , 2017, AAAI.

[13]  A. Tsai,et al.  Nonequilibrium phase diagrams of ternary amorphous alloys , 1997 .

[14]  John J. Irwin,et al.  ZINC 15 – Ligand Discovery for Everyone , 2015, J. Chem. Inf. Model..

[15]  Amin Saberi,et al.  Stochastic Submodular Maximization , 2008, WINE.

[16]  Roman Garnett,et al.  Bayesian Optimal Active Search and Surveying , 2012, ICML.

[17]  Xin Wen,et al.  BindingDB: a web-accessible database of experimentally determined protein–ligand binding affinities , 2006, Nucleic Acids Res..