Machine Learning Uncovers Food- and Excipient-Drug Interactions

Daniel Reker,Yunhua Shi,Ameya R Kirtane,Kaitlyn Hess,Grace J Zhong,Evan Crane,Chih-Hsin Lin,Robert Langer,Giovanni Traverso

doi:10.1016/j.celrep.2020.02.094

Abstract

SUMMARYInactive ingredients and generally recognized as safe compounds are regarded by the US Food and Drug Administration (FDA) as benign for human consumption within specified dose ranges, but a growing body of research has revealed that many inactive ingredients might have unknown biological effects at these concentrations and might alter treatment outcomes. To speed up such discoveries, we apply state-of-the-art machine learning to delineate currently unknown biological effects of inactive ingredients—focusing on P-glycoprotein (P-gp) and uridine diphosphate-glucuronosyltransferase-2B7 (UGT2B7), two proteins that impact the pharmacokinetics of approximately 20% of FDA-approved drugs. Our platform identifies vitamin A palmitate and abietic acid as inhibitors of P-gp and UGT2B7, respectively; in silico, in vitro, ex vivo, and in vivo validations support these interactions. Our predictive framework can elucidate biological effects of commonly consumed chemical matter with implications on food-and excipient-drug interactions and functional drug formulation development.

Highlights

Generally recognized as safe (GRAS)/inactive ingredients (IIGs) compounds could have beneficial biological effects that might be currently underappreciated (Martinez-Mayorga et al, 2013)
Cell Reports 30, 3710–3716, March 17, 2020 3711 descriptions of chemical substructures, we observed a substantial overlap between GRAS/IIG compounds and approved drugs (Figure 1C)
Many of the GRAS/IIG compounds have been previously measured in functional or phenotypic assays and can elicit relevant biological activity: a total of 877 positive assay readouts have been confirmed for GRAS and IIG compounds according to ChEMBL22 data (Figure 1D), which compiles data both from the literature as well as from larger screening efforts (Bento et al, 2014)

Summary

Introduction

GRAS/IIG compounds could have beneficial biological effects that might be currently underappreciated (Martinez-Mayorga et al, 2013) These could provide prime starting points for drug discovery and as functional foods (Martinez-Mayorga and Medina-Franco, 2014), given the well-understood safety, metabolism, and pharmacokinetics of such compounds (Burdock and Carabin, 2004). They might warrant the rational design of functional formulations, which will enable the translation of therapeutics to patients that are currently restricted through unfavorable liberation, absorption, distribution, metabolism, excretion, and toxicity (LADMET) profiles. We use state-of-theart machine learning to predict biologic targets of GRAS/IIG compounds to gain further insights into the biological effects of these essential compound classes and provide innovative starting points for drug discovery and drug formulation research

Methods

Results

Discussion

Conclusion