1
|
Rusic D, Kumric M, Seselja Perisin A, Leskur D, Bukic J, Modun D, Vilovic M, Vrdoljak J, Martinovic D, Grahovac M, Bozic J. Tackling the Antimicrobial Resistance "Pandemic" with Machine Learning Tools: A Summary of Available Evidence. Microorganisms 2024; 12:842. [PMID: 38792673 PMCID: PMC11123121 DOI: 10.3390/microorganisms12050842] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Key Words] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 03/16/2024] [Revised: 04/16/2024] [Accepted: 04/19/2024] [Indexed: 05/26/2024] Open
Abstract
Antimicrobial resistance is recognised as one of the top threats healthcare is bound to face in the future. There have been various attempts to preserve the efficacy of existing antimicrobials, develop new and efficient antimicrobials, manage infections with multi-drug resistant strains, and improve patient outcomes, resulting in a growing mass of routinely available data, including electronic health records and microbiological information that can be employed to develop individualised antimicrobial stewardship. Machine learning methods have been developed to predict antimicrobial resistance from whole-genome sequencing data, forecast medication susceptibility, recognise epidemic patterns for surveillance purposes, or propose new antibacterial treatments and accelerate scientific discovery. Unfortunately, there is an evident gap between the number of machine learning applications in science and the effective implementation of these systems. This narrative review highlights some of the outstanding opportunities that machine learning offers when applied in research related to antimicrobial resistance. In the future, machine learning tools may prove to be superbugs' kryptonite. This review aims to provide an overview of available publications to aid researchers that are looking to expand their work with new approaches and to acquaint them with the current application of machine learning techniques in this field.
Collapse
Affiliation(s)
- Doris Rusic
- Department of Pharmacy, University of Split School of Medicine, Soltanska 2A, 21000 Split, Croatia; (D.R.); (A.S.P.); (D.L.); (J.B.); (D.M.)
| | - Marko Kumric
- Department of Pathophysiology, University of Split School of Medicine, Soltanska 2A, 21000 Split, Croatia; (M.K.); (M.V.); (J.V.); (D.M.)
- Laboratory for Cardiometabolic Research, University of Split School of Medicine, Soltanska 2A, 21000 Split, Croatia
| | - Ana Seselja Perisin
- Department of Pharmacy, University of Split School of Medicine, Soltanska 2A, 21000 Split, Croatia; (D.R.); (A.S.P.); (D.L.); (J.B.); (D.M.)
| | - Dario Leskur
- Department of Pharmacy, University of Split School of Medicine, Soltanska 2A, 21000 Split, Croatia; (D.R.); (A.S.P.); (D.L.); (J.B.); (D.M.)
| | - Josipa Bukic
- Department of Pharmacy, University of Split School of Medicine, Soltanska 2A, 21000 Split, Croatia; (D.R.); (A.S.P.); (D.L.); (J.B.); (D.M.)
| | - Darko Modun
- Department of Pharmacy, University of Split School of Medicine, Soltanska 2A, 21000 Split, Croatia; (D.R.); (A.S.P.); (D.L.); (J.B.); (D.M.)
| | - Marino Vilovic
- Department of Pathophysiology, University of Split School of Medicine, Soltanska 2A, 21000 Split, Croatia; (M.K.); (M.V.); (J.V.); (D.M.)
- Laboratory for Cardiometabolic Research, University of Split School of Medicine, Soltanska 2A, 21000 Split, Croatia
| | - Josip Vrdoljak
- Department of Pathophysiology, University of Split School of Medicine, Soltanska 2A, 21000 Split, Croatia; (M.K.); (M.V.); (J.V.); (D.M.)
- Laboratory for Cardiometabolic Research, University of Split School of Medicine, Soltanska 2A, 21000 Split, Croatia
| | - Dinko Martinovic
- Department of Pathophysiology, University of Split School of Medicine, Soltanska 2A, 21000 Split, Croatia; (M.K.); (M.V.); (J.V.); (D.M.)
- Department of Maxillofacial Surgery, University Hospital of Split, Spinciceva 1, 21000 Split, Croatia
| | - Marko Grahovac
- Department of Pharmacology, University of Split School of Medicine, Soltanska 2A, 21000 Split, Croatia;
| | - Josko Bozic
- Department of Pathophysiology, University of Split School of Medicine, Soltanska 2A, 21000 Split, Croatia; (M.K.); (M.V.); (J.V.); (D.M.)
- Laboratory for Cardiometabolic Research, University of Split School of Medicine, Soltanska 2A, 21000 Split, Croatia
| |
Collapse
|
2
|
Harris PNA, Bauer MJ, Lüftinger L, Beisken S, Forde BM, Balch R, Cotta M, Schlapbach L, Raman S, Shekar K, Kruger P, Lipman J, Bialasiewicz S, Coin L, Roberts JA, Paterson DL, Irwin AD. Rapid nanopore sequencing and predictive susceptibility testing of positive blood cultures from intensive care patients with sepsis. Microbiol Spectr 2024; 12:e0306523. [PMID: 38193658 PMCID: PMC10846127 DOI: 10.1128/spectrum.03065-23] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Key Words] [MESH Headings] [Grants] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 09/01/2023] [Accepted: 12/15/2023] [Indexed: 01/10/2024] Open
Abstract
We aimed to evaluate the performance of Oxford Nanopore Technologies (ONT) sequencing from positive blood culture (BC) broths for bacterial identification and antimicrobial susceptibility prediction. Patients with suspected sepsis in four intensive care units were prospectively enrolled. Human-depleted DNA was extracted from positive BC broths and sequenced using ONT (MinION). Species abundance was estimated using Kraken2, and a cloud-based system (AREScloud) provided in silico predictive antimicrobial susceptibility testing (AST) from assembled contigs. Results were compared to conventional identification and phenotypic AST. Species-level agreement between conventional methods and AST predicted from sequencing was 94.2% (49/52), increasing to 100% in monomicrobial infections. In 262 high-quality AREScloud AST predictions across 24 samples, categorical agreement (CA) was 89.3%, with major error (ME) and very major error (VME) rates of 10.5% and 12.1%, respectively. Over 90% CA was achieved for some taxa (e.g., Staphylococcus aureus) but was suboptimal for Pseudomonas aeruginosa. In 470 AST predictions across 42 samples, with both high quality and exploratory-only predictions, overall CA, ME, and VME rates were 87.7%, 8.3%, and 28.4%. VME rates were inflated by false susceptibility calls in a small number of species/antibiotic combinations with few representative resistant isolates. Time to reporting from sequencing could be achieved within 8-16 h from BC positivity. Direct sequencing from positive BC broths is feasible and can provide accurate predictive AST for some species. ONT-based approaches may be faster but significant improvements in accuracy are required before it can be considered for clinical use.IMPORTANCESepsis and bloodstream infections carry a high risk of morbidity and mortality. Rapid identification and susceptibility prediction of causative pathogens, using Nanopore sequencing direct from blood cultures, may offer clinical benefit. We assessed this approach in comparison to conventional phenotypic methods and determined the accuracy of species identification and susceptibility prediction from genomic data. While this workflow holds promise, and performed well for some common bacterial species, improvements in sequencing accuracy and more robust predictive algorithms across a diverse range of organisms are required before this can be considered for clinical use. However, results could be achieved in timeframes that are faster than conventional phenotypic methods.
Collapse
Affiliation(s)
- Patrick N. A. Harris
- UQ Centre for Clinical Research, Faculty of Medicine, University of Queensland, Brisbane, Australia
- Central Microbiology, Pathology Queensland, Royal Brisbane and Women’s Hospital, Brisbane, Australia
- Herston Infectious Disease Institute, Royal Brisbane and Women’s Hospital Campus, Brisbane, Australia
| | - Michelle J. Bauer
- UQ Centre for Clinical Research, Faculty of Medicine, University of Queensland, Brisbane, Australia
| | | | | | - Brian M. Forde
- UQ Centre for Clinical Research, Faculty of Medicine, University of Queensland, Brisbane, Australia
| | - Ross Balch
- UQ Centre for Clinical Research, Faculty of Medicine, University of Queensland, Brisbane, Australia
| | - Menino Cotta
- UQ Centre for Clinical Research, Faculty of Medicine, University of Queensland, Brisbane, Australia
| | - Luregn Schlapbach
- University Children’s Hospital Zurich, University of Zurich, Zurich, Switzerland
- Child Health Research Centre, The University of Queensland, Brisbane, Queensland, Australia
| | - Sainath Raman
- Child Health Research Centre, The University of Queensland, Brisbane, Queensland, Australia
- Paediatric Intensive Care Unit, Queensland Children’s Hospital, South Brisbane, Australia
| | - Kiran Shekar
- Adult Intensive Care Services, The Prince Charles Hospital, Brisbane, Queensland, Australia
- Faculty of Medicine, University of Queensland, Brisbane, Australia
| | - Peter Kruger
- Intensive Care Unit, Princess Alexandra Hospital, Woolloongabba, Queensland, Australia
- Department of Anaesthesiology and Critical Care, The University of Queensland, St Lucia, Queensland, Australia
| | - Jeff Lipman
- UQ Centre for Clinical Research, Faculty of Medicine, University of Queensland, Brisbane, Australia
- Intensive Care Unit, Royal Brisbane and Women’s Hospital, Brisbane, Australia
- Division of Anaesthesiology Critical Care Emergency and Pain Medicine, Nîmes University Hospital, University of Montpellier, Nîmes, France
- Jamieson Trauma Institute, Royal Brisbane and Women’s Hospital, Brisbane, Australia
| | - Seweryn Bialasiewicz
- Australian Centre for Ecogenomics, School of Chemistry and Molecular Biosciences, Faculty of Science, University of Queensland, Brisbane, Australia
| | - Lachlan Coin
- Department of Microbiology and Immunology, The Peter Doherty Institute for Infection and Immunity, University of Melbourne, Melbourne, Victoria, Australia
| | - Jason A. Roberts
- UQ Centre for Clinical Research, Faculty of Medicine, University of Queensland, Brisbane, Australia
- Herston Infectious Disease Institute, Royal Brisbane and Women’s Hospital Campus, Brisbane, Australia
- Division of Anaesthesiology Critical Care Emergency and Pain Medicine, Nîmes University Hospital, University of Montpellier, Nîmes, France
- Departments of Pharmacy and Intensive Care Medicine, Royal Brisbane and Women’s Hospital, Brisbane, Australia
| | - David L. Paterson
- UQ Centre for Clinical Research, Faculty of Medicine, University of Queensland, Brisbane, Australia
- ADVANCE-ID, Saw Swee School of Public Health, National University of Singapore, Singapore, Singapore
| | - Adam D. Irwin
- UQ Centre for Clinical Research, Faculty of Medicine, University of Queensland, Brisbane, Australia
- Infection Management and Prevention Service, Queensland Children’s Hospital, Brisbane, Queensland, Australia
| |
Collapse
|
3
|
Hanafiah A, Sukri A, Yusoff H, Chan CS, Hazrin-Chong NH, Salleh SA, Neoh HM. Insights into the Microbiome and Antibiotic Resistance Genes from Hospital Environmental Surfaces: A Prime Source of Antimicrobial Resistance. Antibiotics (Basel) 2024; 13:127. [PMID: 38391513 PMCID: PMC10885873 DOI: 10.3390/antibiotics13020127] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Key Words] [Grants] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 12/21/2023] [Revised: 01/20/2024] [Accepted: 01/25/2024] [Indexed: 02/24/2024] Open
Abstract
Hospital environmental surfaces are potential reservoirs for transmitting hospital-associated pathogens. This study aimed to profile microbiomes and antibiotic resistance genes (ARGs) from hospital environmental surfaces using 16S rRNA amplicon and metagenomic sequencing at a tertiary teaching hospital in Malaysia. Samples were collected from patient sinks and healthcare staff counters at surgery and orthopaedic wards. The samples' DNA were subjected to 16S rRNA amplicon and shotgun sequencing to identify bacterial taxonomic profiles, antibiotic resistance genes, and virulence factor pathways. The bacterial richness was more diverse in the samples collected from patient sinks than those collected from staff counters. Proteobacteria and Verrucomicrobia dominated at the phylum level, while Bacillus, Staphylococcus, Pseudomonas, and Acinetobacter dominated at the genus level. Staphylococcus epidermidis and Staphylococcus aureus were prevalent on sinks while Bacillus cereus dominated the counter samples. The highest counts of ARGs to beta-lactam were detected, followed by ARGs against fosfomycin and cephalosporin. We report the detection of mcr-10.1 that confers resistance to colistin at a hospital setting in Malaysia. The virulence gene pathways that aid in antibiotic resistance gene transfer between bacteria were identified. Environmental surfaces serve as potential reservoirs for nosocomial infections and require mitigation strategies to control the spread of antibiotic resistance bacteria.
Collapse
Affiliation(s)
- Alfizah Hanafiah
- Department of Medical Microbiology and Immunology, Faculty of Medicine, Universiti Kebangsaan Malaysia, Cheras, Kuala Lumpur 56000, Malaysia
| | - Asif Sukri
- Department of Biological Sciences and Biotechnology, Faculty of Science and Technology, Universiti Kebangsaan Malaysia, Bangi 43600, Malaysia
| | - Hamidah Yusoff
- Department of Medical Microbiology and Immunology, Faculty of Medicine, Universiti Kebangsaan Malaysia, Cheras, Kuala Lumpur 56000, Malaysia
| | | | - Nur Hazlin Hazrin-Chong
- Department of Biological Sciences and Biotechnology, Faculty of Science and Technology, Universiti Kebangsaan Malaysia, Bangi 43600, Malaysia
| | - Sharifah Azura Salleh
- Infection Control Unit, Hospital Canselor Tuanku Muhriz, Cheras, Kuala Lumpur 56000, Malaysia
| | - Hui-Min Neoh
- UKM Molecular Biology Institute (UMBI), Universiti Kebangsaan Malaysia, Cheras, Kuala Lumpur 56000, Malaysia
| |
Collapse
|