A Deep Learning Approach to Antibiotic Discovery

  title={A Deep Learning Approach to Antibiotic Discovery},
  author={Jonathan M. Stokes and Kevin Yang and Kyle Swanson and Wengong Jin and Andres Cubillos-Ruiz and Nina M. Donghia and Craig R MacNair and Shawn French and Lindsey A. Carfrae and Zohar Bloom-Ackermann and Victoria M. Tran and Anush Chiappino-Pepe and Ahmed H. Badran and Ian W. Andrews and Emma J. Chory and George M. Church and Eric D. Brown and T. Jaakkola and Regina Barzilay and James J. Collins},

Figures from this paper

A brief guide to machine learning for antibiotic discovery.

Accelerated antimicrobial discovery via deep generative models and molecular dynamics simulations

Two antimicrobial peptides have low toxicity, as validated in vitro and in mice, and it is shown using live-cell confocal imaging that the bactericidal mode of action of the peptides involves the formation of membrane pores.

An interpretable machine learning approach to identify mechanism of action of antibiotics

InterPred is introduced, an interpretable technique for predicting bioactivity of small molecules and their mechanism of action that has the same accuracy as the state of the art in bioactivity prediction, and it enables assigning chemical moieties that are responsible for bioactivity.

Predicting Antimicrobial Activity of Conjugated Oligoelectrolyte Molecules via Machine Learning.

An antimicrobial activity prediction model for conjugated oligoelectrolyte molecules, a new class of antibiotics that lacks extensive prior structure-activity relationship studies, is developed and can be readily adapted to other novel antibiotic candidate domains.

Accelerating antibiotic discovery through artificial intelligence

This review describes AI-facilitated advances in the discovery of both small molecule antibiotics and antimicrobial peptides and analyzes uptake of open science best practices in AI-driven antibiotic discovery and argues for openness and reproducibility as a means of accelerating preclinical research.

Identification of antimicrobial peptides from the human gut microbiome using deep learning.

This study combined multiple natural language processing neural network models, including LSTM, Attention and BERT, to form a unified pipeline for candidate AMP identification from human gut microbiome data and demonstrated significant efficacy in lowering bacterial load against a mouse model of bacterial lung infection.

Rational Discovery of Antimicrobial Peptides by Means of Artificial Intelligence

This work proposes implementing an artificial intelligence algorithm to reduce the required experimentation and increase the efficiency of high-activity AMP discovery, and validates the previously proposed pipeline for a more efficient rational discovery of novel AMPs.

Parsing Molecules for Drug Discovery

Chemprop’s ability to increase the efficiency of screens for antibiotics by applying it to three different compound libraries, the Drug Repurposing Hub, the WuXi antituberculosis, and the ZINC15 libraries is demonstrated.

Deep Neural Networks in the Discovery of Novel Antibiotics Drug Molecule: A Review

The use of the artificial intelligence techniques in the antibiotics discovery will be reviewed with focus on the deep neural networks model as compared with other methods.

From Genome to Drugs: New Approaches in Antimicrobial Discovery

This review focuses on how the advent of omics and the development and application of bioinformatics strategies conduct a “big-data era” that improves target selection and lead compound identification in a cost-effective and shortened timeline.



A Common Platform for Antibiotic Dereplication and Adjuvant Discovery.

ESKAPEing the labyrinth of antibacterial discovery

Findings from target-based and phenotypic screening efforts carried out at AstraZeneca over the past decade are described, some of the subsequent chemistry challenges are discussed and a description of new approaches comprising a combination of computational modelling and advanced biological tools which may pave the way towards the discovery of new antibacterial agents are described.

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

Unconventional screening approaches for antibiotic discovery

Recent nontraditional approaches that have been piloted for early drug discovery efforts open new avenues for finding the next generation of antimicrobials.

Opportunities for natural products in 21st century antibiotic discovery.

21st century natural product research is perfectly positioned to fill the antibiotic discovery gap and bring new drug candidates to the clinic and is poised to regain prominence in delivering new drugs to solve the antibiotic crisis.

Antibacterial drug discovery in the resistance era

The looming antibiotic-resistance crisis has penetrated the consciousness of clinicians, researchers, policymakers, politicians and the public at large. The evolution and widespread distribution of

A multiplexable assay for screening antibiotic lethality against drug-tolerant bacteria

The solid media portable cell killing assay uses metabolism-sensitive staining to illuminate the killing of antibiotic-tolerant bacteria under resource-depleted conditions, thereby enabling multiplex, genome-scale analyses for the identification of target strains.

How antibiotics kill bacteria: from targets to networks

The multilayered effects of drug–target interactions, including the essential cellular processes that are inhibited by bactericidal antibiotics and the associated cellular response mechanisms that contribute to killing are discussed.

Pentamidine sensitizes Gram-negative pathogens to antibiotics and overcomes acquired colistin resistance

An unconventional screening platform designed to enrich for non-lethal, outer-membrane-active compounds with potential as adjuvants for conventional antibiotics is described, identifying the antiprotozoal drug pentamidine6 as an effective perturbant of the Gram-negative outer membrane through its interaction with lipopolysaccharide.