Project

Back to overview

PERPHECT: Deep Generative Networks for Bacteriophage Genetic Edition

English title PERPHECT: Deep Generative Networks for Bacteriophage Genetic Edition
Applicant Pena-Reyes Carlos Andrés
Number 190298
Funding scheme Spark
Research institution Haute Ecole d'Ingénierie et de Gestion du Canton de Vaud (HEIG-VD)
Institution of higher education University of Applied Sciences and Arts Western Switzerland - HES-SO
Main discipline Information Technology
Start/End 01.12.2019 - 31.10.2021
Approved amount 87'240.00
Show all

All Disciplines (2)

Discipline
Information Technology
Genetics

Keywords (5)

Machine Learning; Deep Learning; Deep-generative networks; genetically-engineered phages; Phage therapy

Lay Summary (French)

Lead
La résistance bactérienne aux antibiotiques est devenue un phénomène global préoccupant pouvant amener à une impasse thérapeutique. Cette situation a suscité l'intérêt de la phagothérapie, une alternative thérapeutique qui utilise des virus prédateurs naturels des bactéries appelées « bactériophages ». Un bactériophage est très spécifique, il ne tuera qu’une seule souche bactérienne ce qui en fait un outil thérapeutique extrêmement précis. Cependant l’utilisation peut être entravée par la difficulté à trouver le ou les phages adéquats, d’autant plus qu’il faut administrer une combinaison de phages pour s'assurer de l’efficacité du traitement. Pour résoudre cette difficulté, le génie génétique permet de modifier le génome des phages, leur octroyant la capacité de détruire une bactérie cible. Cependant une des limites de cette approche réside dans sa forte dépendance de connaissances empiriques pour guider les modifications génétiques.
Lay summary

Notre principal objectif est d’appliquer l'intelligence artificielle (IA) pour à la fois guider l'ingénierie génomique des phages, augmenter la vitesse de modification des génomes et, in-fine, à améliorer l'activité des phages. Avec PERPHECT, nous visons à explorer et à évaluer le potentiel de méthodes capables de générer des séquences ADN et cibler les modifications génétiques permettant d’optimiser l'interaction phage-bactérie. Le projet se concentre sur l’utilisation d’un modèle génératif où deux réseaux de neurones artificiels sont placés en compétition. Le premier, le générateur, produit une séquence génétique très similaire à celles qui existent dans la nature. Le second, son adversaire, le discriminateur prédit les interactions phage-bactérie à partir des séquences génétiques générées. Il essaie de détecter si la séquence est proche de la réalité ou bien si elle est artificielle et sans valeur biologique. Les séquences générées par cette approche visent à maximiser l'interaction des phages avec une bactérie cible, tout en minimisant les effets potentiellement perturbateurs.

Ce projet permettra de mieux comprendre l’organisation des génomes de phage, de déterminer les zones d’intérêts, et d’identifier les mécanismes clés dans les interactions phages-bactéries. La capacité d’explorer les génomes et de guider l'ingénierie génomique des phages par une approche basée sur l’IA permettra de générer ou modifier son génome en supprimant, en introduisant, ou en remplaçant une partie de son ADN pour obtenir des phages efficaces contre une bactérie cible. Ces méthodologies ouvriront la voie à une stratégie antibactérienne efficiente et adaptable, qui aura un impact essentiel sur le traitement futur des infections bactériennes et sur le développement de l’ingénierie génétique.

Direct link to Lay Summary Last update: 18.12.2019

Responsible applicant and co-applicants

Employees

Publications

Publication
Towards BacterioPhage Genetic Edition: Deep Learning Prediction of Phage-Bacterium Interactions
Ataee Shabnam, Rodriguez Oscar, Brochet Xavier, Pena Carlos Andres (2020), Towards BacterioPhage Genetic Edition: Deep Learning Prediction of Phage-Bacterium Interactions, in 2020 IEEE International Conference on Bioinformatics and Biomedicine (BIBM), Seoul, Korea (South)IEEE, USA.

Collaboration

Group / person Country
Types of collaboration
Head of the laboratory of bacteriophages and phage therapy/Center for Research and Innovation in Cli Switzerland (Europe)
- in-depth/constructive exchanges on approaches, methods or results

Scientific events

Active participation

Title Type of contribution Title of article or contribution Date Place Persons involved
SIB Course on “Deep Learning for Life Sciences” Individual talk Deep Learning Prediction of Phage-Bacterium Interactions & Towards Bacteriophage Genetic Editing 25.11.2021 Streamed, Switzerland Pena-Reyes Carlos Andrés;
Machine Learning in Swiss bioinformatics: applications and challenges Individual talk Talking trends in bioinformatics 07.12.2020 Virtual , Switzerland Pena-Reyes Carlos Andrés;
SIB Days 2020 Talk given at a conference PERPHECT: Deep Generative Networks for BacterioPhage Genetic Edition 08.06.2020 virtual conference, Switzerland Pena-Reyes Carlos Andrés;


Communication with the public

Communication Title Media Place Year
Talks/events/exhibitions Artificial Intelligence to tackle bacteria, viruses and other bugs International 2021

Abstract

Bacterial infections will become highly life-threatening events if nothing is done to curb the current trend of antimicrobial resistance. The paucity of potential new anti-infectives in the pipeline of pharmaceutical industries has called for the development of new treatment strategies and incited a renewed interest in bacteriophage (phage) therapy, i.e., the direct use of natural, microbial viruses for the treatment of bacterial infections, as such a potential alternative.The close evolutionary relationship between phages and their bacterial hosts entails that their genetic information can be used to predict their interactions. The advent of high-throughput experimental techniques paved the way to genome-wide computational analysis and predictive studies. This has motivated several attempts to use bacterial and phage genome analysis, combined with data about known interactions, to train computational models that can predict interactions between new unknown bacteria and/or phages. Phage therapy efficacy may be hindered by the narrow host range and very high specificity of phages which causes (i) that finding the appropriate phages is a difficult task and (ii) obliging to produce and use cocktails of several phages to fight against simple infections. To solve this difficulty, one forward-thinking modernization of phage therapy is to use genetically-engineered (GE) phages. Such engineered viruses can provide substantial advantages over natural phage in terms of host range, immune system recognition, and environmental stability. A limitation for GE-phages is the empirical knowledge that is currently used to guide genetic modifications. Phages are estimated to be the most abundant organisms on Earth with unparalleled genetic diversity, making it impossible for basic research techniques to factor in all possible variables for creating the most active GE phages.New technologies are further needed to accelerate the design-build-test cycle for engineering phages and to make it possible to translate proof-of-concept academic work into real-world use more efficiently. The application of Artificial intelligence (AI) to this context has potential both to increase the speed at which genomes can be engineered and to enhance the activity of resulting phages. The only applications of AI to phage biology concern mainly predicting species- and/or strain-level interactions, host species prediction, or the identification of novel viral sequences, making the idea proposed in this project highly innovative. This project, PERPHECT, focuses on the combined use of Deep Neural Networks, that enable extracting patterns directly from DNA sequences which are relevant to predict phage-bacteria interactions, and Deep Generative Adversarial Networks, that have the potential to create sequences, very similar to naturally occurring ones. The sequences generated with this approach are intended to maximize phage interaction with a target bacterium, while minimizing potentially-perturbing, length-associated effects such as noise, repeats, and non-informative code. Such essential segments are, then, integrated into selected, existing, phage genomes thanks to a genome-editing strategy based on sequence similarity and alignment. Finally, the resulting engineered phage genomes are evaluated in silico by means of several interaction-prediction computational models so as to validate their potential interaction.
-