Applying support-vector machine learning algorithms toward predicting host-guest interactions with cucurbit[7]uril.

Anthony Tabet,Guanglu Wu,Polina Anikeeva,Harry Bulstrode,Merrick Pierson Smela,Charlie Readman,Vijay K Rana,David H Rowitch,Oren A Scherman,Thomas Gebhart,Cole Baker

doi:10.1039/c9cp05800a

Abstract

Machine learning is a valuable tool in the development of chemical technologies but its applications into supramolecular chemistry have been limited. Here, the utility of kernel-based support vector machine learning using density functional theory calculations as training data is evaluated when used to predict equilibrium binding coefficients of small molecules with cucurbit[7]uril (CB[7]). We find that utilising SVMs may confer some predictive ability. This algorithm was then used to predict the binding of drugs TAK-580 and selumetinib. The algorithm did predict strong binding for TAK-580 and poor binding for selumetinib, and these results were experimentally validated. It was discovered that the larger homologue cucurbit[8]uril (CB[8]) is partial to selumetinib, suggesting an opportunity for tunable release by introducing different concentrations of CB[7] or CB[8] into a hydrogel depot. We qualitatively demonstrated that these drugs may have utility in combination against gliomas. Finally, mass transfer simulations show CB[7] can independently tune the release of TAK-580 without affecting selumetinib. This work gives specific evidence that a machine learning approach to recognition of small molecules by macrocycles has merit and reinforces the view that machine learning may prove valuable in the development of drug delivery systems and supramolecular chemistry more broadly.

Highlights

IntroductionThe applications of machine learning in biology and chemistry have rapidly expanded in recent years due to the potential of data science to improve small molecule drug discovery, identify more efficient synthetic pathways, create proteins with greater binding affinity to specific substrates, and other applications. One application that has not yet been explored is predicting the molecular recognition of small molecules with macrocyclic hosts
The applications of machine learning in biology and chemistry have rapidly expanded in recent years due to the potential of data science to improve small molecule drug discovery, identify more efficient synthetic pathways, create proteins with greater binding affinity to specific substrates, and other applications.1–5 One application that has not yet been explored is predicting the molecular recognition of small molecules with macrocyclic hosts.c Department of Materials Science & Engineering and Research Laboratory of Electronics, Massachusetts Institute of Technology, Cambridge, MA, USA d Department of Computer Science, University of Minnesota, Minneapolis, MN, USA e Department of Chemistry and Chemical Biology, Harvard University, Cambridge, MA, USA f Department of Electrical Engineering & Computer Science, Massachusetts Institute of Technology, Cambridge, MA, USA † Electronic supplementary information (ESI) available
Machine learning is a valuable tool in the development of chemical technologies but its applications into supramolecular chemistry have been limited

Summary

Introduction

The applications of machine learning in biology and chemistry have rapidly expanded in recent years due to the potential of data science to improve small molecule drug discovery, identify more efficient synthetic pathways, create proteins with greater binding affinity to specific substrates, and other applications. One application that has not yet been explored is predicting the molecular recognition of small molecules with macrocyclic hosts. Cucurbiturils are a class of symmetric macrocycles that have applications within drug delivery, biosensing, catalysis, and energy.. Cucurbiturils are a class of symmetric macrocycles that have applications within drug delivery, biosensing, catalysis, and energy.6,7 These macrocycles have many advantages over their non-symmetric counterparts such as cyclodextrins, including temperature stability and robustness at acidic and basic pH values, such as those that occur naturally in physiology. Cucurbituril acts as a competitive substrate and binds to the active ingredient This binding can reduce the effective concentration and increase the half life of biologic and hydrophobic small molecule drugs.. We report the utility of this regression in predicting the binding of two new small molecule drugs that have received promising. We provide a qualitative example of the potential use of these predictions in developing cocktail drug therapies against a pediatric low grade glioma cell model

Methods and model

Results and discussion

Conclusion