scholarly journals Direct antimicrobial resistance prediction from clinical MALDI-TOF mass spectra using machine learning

2022 ◽  
Author(s):  
Caroline Weis ◽  
Aline Cuénod ◽  
Bastian Rieck ◽  
Olivier Dubuis ◽  
Susanne Graf ◽  
...  
2020 ◽  
Author(s):  
Caroline Weis ◽  
Aline Cuénod ◽  
Bastian Rieck ◽  
Felipe Llinares-López ◽  
Olivier Dubuis ◽  
...  

AbstractEarly administration of effective antimicrobial treatments improves the outcome of infections. Culture-based antimicrobial resistance testing allows for tailored treatments, but takes up to 96h. We present a revolutionary approach to predict resistance with unmatched speed within 24h, using calibrated logistic regression and LightGBM-classifiers trained on species-specific MALDI-TOF mass spectrometry measurements. For this analysis, we created an unprecedented large, publicly-available dataset combining mass spectra and resistance information. Our models provide highly valuable treatment guidance 12–72h earlier than classical approaches. Rejection of uncertain predictions enables quality control and clinically-applicable sensitivities and specificities for the priority pathogens Staphylococcus aureus, Escherichia coli, and Klebsiella pneumoniae.


2011 ◽  
Vol 34 (1) ◽  
pp. 20-29 ◽  
Author(s):  
Katrien De Bruyne ◽  
Bram Slabbinck ◽  
Willem Waegeman ◽  
Paul Vauterin ◽  
Bernard De Baets ◽  
...  

2021 ◽  
Vol 9 (2) ◽  
pp. 416
Author(s):  
Charles Dumolin ◽  
Charlotte Peeters ◽  
Evelien De Canck ◽  
Nico Boon ◽  
Peter Vandamme

Culturomics-based bacterial diversity studies benefit from the implementation of MALDI-TOF MS to remove genomically redundant isolates from isolate collections. We previously introduced SPeDE, a novel tool designed to dereplicate spectral datasets at an infraspecific level into operational isolation units (OIUs) based on unique spectral features. However, biological and technical variation may result in methodology-induced differences in MALDI-TOF mass spectra and hence provoke the detection of genomically redundant OIUs. In the present study, we used three datasets to analyze to which extent hierarchical clustering and network analysis allowed to eliminate redundant OIUs obtained through biological and technical sample variation and to describe the diversity within a set of spectra obtained from 134 unknown soil isolates. Overall, network analysis based on unique spectral features in MALDI-TOF mass spectra enabled a superior selection of genomically diverse OIUs compared to hierarchical clustering analysis and provided a better understanding of the inter-OIU relationships.


2017 ◽  
Vol 53 (2) ◽  
pp. 162-171 ◽  
Author(s):  
Andrea R. Kelley ◽  
Madeline E. Colley ◽  
George Perry ◽  
Stephan B.H. Bach

2007 ◽  
Vol 79 (4) ◽  
pp. 1639-1645 ◽  
Author(s):  
Alena Krupková ◽  
Jan Čermák ◽  
Zuzana Walterová ◽  
Jiří Horský

1999 ◽  
Vol 71 (15) ◽  
pp. 3226-3230 ◽  
Author(s):  
Ricky D. Holland ◽  
Christopher R. Duffy ◽  
Fatemeh Rafii ◽  
John B. Sutherland ◽  
Thomas M. Heinze ◽  
...  

2007 ◽  
Vol 24 (1) ◽  
pp. 63-70 ◽  
Author(s):  
D. Mantini ◽  
F. Petrucci ◽  
P. Del Boccio ◽  
D. Pieragostino ◽  
M. Di Nicola ◽  
...  

2011 ◽  
Vol 76 (12) ◽  
pp. 1687-1701 ◽  
Author(s):  
Bojana Damnjanovic ◽  
Biljana Petrovic ◽  
Jasmina Dimitric-Markovic ◽  
Marijana Petkovic

In this work, the matrix-assisted laser desorption and ionization time-of-flight (MALDI-TOF) mass spectra of two cationic complexes, i.e., [PdCl(dien)]Cl and [Ru(en)2Cl2]Cl, acquired under different conditions were analyzed. The spectra were recorded with three matrices with or without trifluoroacetic acid (TFA), i.e., two traditional matrices, i.e., 2,5-dihydroxybenzoic acid and ?-cyano-hydroxycinnamic acid, and one flavonoid, quercetin. The spectra acquired with quercetin appeared to be the simplest, whereas in the spectra obtained with other matrices, peaks arising either from the addition of matrix molecules or from the fragmentation products were detectable. Addition of TFA did not complicate the spectra of the Pd(II) and Ru(III) complexes when the traditional matrices were used. On the other hand, the spectra of Pd complex were simpler, whereas the addition of TFA in the case of the Ru complex resulted in a higher number of peaks, some of which could not be identified. Taken together, the results of this study once more emphasize the differences arising in the MALDI-TOF mass spectra of transition metal complexes in dependence on the applied matrix.


Sign in / Sign up

Export Citation Format

Share Document