A White-Box Machine Learning Approach for Revealing Antibiotic Mechanisms of Action

Jason H Yang, Sarah N Wright, Meagan Hamblin, Douglas McCloskey, Miguel A Alcantar, Lars Schrübbers, Allison J Lopatkin, Sangeeta Satish, Amir Nili, Bernhard O Palsson, Graham C Walker, James J Collins*

*Corresponding author for this work

Research output: Contribution to journalJournal articleResearchpeer-review

Abstract

Current machine learning techniques enable robust association of biological signals with measured phenotypes, but these approaches are incapable of identifying causal relationships. Here, we develop an integrated "white-box" biochemical screening, network modeling, and machine learning approach for revealing causal mechanisms and apply this approach to understanding antibiotic efficacy. We counter-screen diverse metabolites against bactericidal antibiotics in Escherichia coli and simulate their corresponding metabolic states using a genome-scale metabolic network model. Regression of the measured screening data on model simulations reveals that purine biosynthesis participates in antibiotic lethality, which we validate experimentally. We show that antibiotic-induced adenine limitation increases ATP demand, which elevates central carbon metabolism activity and oxygen consumption, enhancing the killing effects of antibiotics. This work demonstrates how prospective network modeling can couple with machine learning to identify complex causal mechanisms underlying drug efficacy.
Original languageEnglish
JournalCELL
Volume177
Issue number6
Pages (from-to)1649-1661.e9
ISSN0092-8674
DOIs
Publication statusPublished - 2019

Keywords

  • Machine learning
  • Network modeling
  • Antibiotics
  • Metabolism
  • Purine biosynthesis
  • ATP
  • Adenylate energy charge
  • NADPH:NADP+ ratio
  • LC-MS/MS
  • Biochemical screen

Cite this

Yang, J. H., Wright, S. N., Hamblin, M., McCloskey, D., Alcantar, M. A., Schrübbers, L., ... Collins, J. J. (2019). A White-Box Machine Learning Approach for Revealing Antibiotic Mechanisms of Action. CELL, 177(6), 1649-1661.e9. https://doi.org/10.1016/j.cell.2019.04.016
Yang, Jason H ; Wright, Sarah N ; Hamblin, Meagan ; McCloskey, Douglas ; Alcantar, Miguel A ; Schrübbers, Lars ; Lopatkin, Allison J ; Satish, Sangeeta ; Nili, Amir ; Palsson, Bernhard O ; Walker, Graham C ; Collins, James J. / A White-Box Machine Learning Approach for Revealing Antibiotic Mechanisms of Action. In: CELL. 2019 ; Vol. 177, No. 6. pp. 1649-1661.e9.
@article{2176e5ec122a42d49161a34e49212815,
title = "A White-Box Machine Learning Approach for Revealing Antibiotic Mechanisms of Action",
abstract = "Current machine learning techniques enable robust association of biological signals with measured phenotypes, but these approaches are incapable of identifying causal relationships. Here, we develop an integrated {"}white-box{"} biochemical screening, network modeling, and machine learning approach for revealing causal mechanisms and apply this approach to understanding antibiotic efficacy. We counter-screen diverse metabolites against bactericidal antibiotics in Escherichia coli and simulate their corresponding metabolic states using a genome-scale metabolic network model. Regression of the measured screening data on model simulations reveals that purine biosynthesis participates in antibiotic lethality, which we validate experimentally. We show that antibiotic-induced adenine limitation increases ATP demand, which elevates central carbon metabolism activity and oxygen consumption, enhancing the killing effects of antibiotics. This work demonstrates how prospective network modeling can couple with machine learning to identify complex causal mechanisms underlying drug efficacy.",
keywords = "Machine learning, Network modeling, Antibiotics, Metabolism, Purine biosynthesis, ATP, Adenylate energy charge, NADPH:NADP+ ratio, LC-MS/MS, Biochemical screen",
author = "Yang, {Jason H} and Wright, {Sarah N} and Meagan Hamblin and Douglas McCloskey and Alcantar, {Miguel A} and Lars Schr{\"u}bbers and Lopatkin, {Allison J} and Sangeeta Satish and Amir Nili and Palsson, {Bernhard O} and Walker, {Graham C} and Collins, {James J}",
year = "2019",
doi = "10.1016/j.cell.2019.04.016",
language = "English",
volume = "177",
pages = "1649--1661.e9",
journal = "Cell",
issn = "0092-8674",
publisher = "Cell Press",
number = "6",

}

Yang, JH, Wright, SN, Hamblin, M, McCloskey, D, Alcantar, MA, Schrübbers, L, Lopatkin, AJ, Satish, S, Nili, A, Palsson, BO, Walker, GC & Collins, JJ 2019, 'A White-Box Machine Learning Approach for Revealing Antibiotic Mechanisms of Action', CELL, vol. 177, no. 6, pp. 1649-1661.e9. https://doi.org/10.1016/j.cell.2019.04.016

A White-Box Machine Learning Approach for Revealing Antibiotic Mechanisms of Action. / Yang, Jason H; Wright, Sarah N; Hamblin, Meagan; McCloskey, Douglas; Alcantar, Miguel A; Schrübbers, Lars; Lopatkin, Allison J; Satish, Sangeeta; Nili, Amir; Palsson, Bernhard O; Walker, Graham C; Collins, James J.

In: CELL, Vol. 177, No. 6, 2019, p. 1649-1661.e9.

Research output: Contribution to journalJournal articleResearchpeer-review

TY - JOUR

T1 - A White-Box Machine Learning Approach for Revealing Antibiotic Mechanisms of Action

AU - Yang, Jason H

AU - Wright, Sarah N

AU - Hamblin, Meagan

AU - McCloskey, Douglas

AU - Alcantar, Miguel A

AU - Schrübbers, Lars

AU - Lopatkin, Allison J

AU - Satish, Sangeeta

AU - Nili, Amir

AU - Palsson, Bernhard O

AU - Walker, Graham C

AU - Collins, James J

PY - 2019

Y1 - 2019

N2 - Current machine learning techniques enable robust association of biological signals with measured phenotypes, but these approaches are incapable of identifying causal relationships. Here, we develop an integrated "white-box" biochemical screening, network modeling, and machine learning approach for revealing causal mechanisms and apply this approach to understanding antibiotic efficacy. We counter-screen diverse metabolites against bactericidal antibiotics in Escherichia coli and simulate their corresponding metabolic states using a genome-scale metabolic network model. Regression of the measured screening data on model simulations reveals that purine biosynthesis participates in antibiotic lethality, which we validate experimentally. We show that antibiotic-induced adenine limitation increases ATP demand, which elevates central carbon metabolism activity and oxygen consumption, enhancing the killing effects of antibiotics. This work demonstrates how prospective network modeling can couple with machine learning to identify complex causal mechanisms underlying drug efficacy.

AB - Current machine learning techniques enable robust association of biological signals with measured phenotypes, but these approaches are incapable of identifying causal relationships. Here, we develop an integrated "white-box" biochemical screening, network modeling, and machine learning approach for revealing causal mechanisms and apply this approach to understanding antibiotic efficacy. We counter-screen diverse metabolites against bactericidal antibiotics in Escherichia coli and simulate their corresponding metabolic states using a genome-scale metabolic network model. Regression of the measured screening data on model simulations reveals that purine biosynthesis participates in antibiotic lethality, which we validate experimentally. We show that antibiotic-induced adenine limitation increases ATP demand, which elevates central carbon metabolism activity and oxygen consumption, enhancing the killing effects of antibiotics. This work demonstrates how prospective network modeling can couple with machine learning to identify complex causal mechanisms underlying drug efficacy.

KW - Machine learning

KW - Network modeling

KW - Antibiotics

KW - Metabolism

KW - Purine biosynthesis

KW - ATP

KW - Adenylate energy charge

KW - NADPH:NADP+ ratio

KW - LC-MS/MS

KW - Biochemical screen

U2 - 10.1016/j.cell.2019.04.016

DO - 10.1016/j.cell.2019.04.016

M3 - Journal article

VL - 177

SP - 1649-1661.e9

JO - Cell

JF - Cell

SN - 0092-8674

IS - 6

ER -