Imperial College London
Portrait Picture

DrLeonidChindelevitch

Faculty of Medicine, School of Public Health

Lecturer in Infectious Disease Epidemiology
 
 
 
//

Contact

 

l.chindelevitch Website

 
 
//

Location

 

Sir Michael Uren HubWhite City Campus

//

Summary

 

Publications

Citation

BibTex format

@article{Zabeti:2020:10.1101/2020.05.31.115741,
author = {Zabeti, H and Dexter, N and Safari, AH and Sedaghat, N and Libbrecht, M and Chindelevitch, L},
doi = {10.1101/2020.05.31.115741},
title = {INGOT-DR: an interpretable classifier for predicting drug resistance in M. tuberculosis},
url = {http://dx.doi.org/10.1101/2020.05.31.115741},
year = {2020}
}
Download

RIS format (EndNote, RefMan)

TY  - JOUR
AB  - <jats:title>Abstract</jats:title><jats:sec><jats:title>Motivation</jats:title><jats:p>Prediction of drug resistance and identification of its mechanisms in bacteria such as <jats:italic>Mycobacterium tuberculosis</jats:italic>, the etiological agent of tuberculosis, is a challenging problem. Solving this problem requires a transparent, accurate, and flexible predictive model. The methods currently used for this purpose rarely satisfy all of these criteria. On the one hand, approaches based on testing strains against a catalogue of previously identified mutations often yield poor predictive performance; on the other hand, machine learning techniques typically have higher predictive accuracy, but often lack interpretability and may learn patterns that produce accurate predictions for the wrong reasons. Current interpretable methods may either exhibit a lower accuracy or lack the flexibility needed to generalize them to previously unseen data.</jats:p></jats:sec><jats:sec><jats:title>Contribution</jats:title><jats:p>In this paper we propose a novel technique, inspired by the group testing and Boolean compressed sensing, which yields highly accurate predictions, interpretable results, and is flexible enough to be optimized for various evaluation metrics at the same time.</jats:p></jats:sec><jats:sec><jats:title>Results</jats:title><jats:p>We test the predictive accuracy of our approach on five first-line and seven second-line antibiotics used for treating tuberculosis. We find that it has a higher or comparable accuracy to that of commonly used machine learning models, and is able to identify variants in genes with previously reported association to drug resistance. Our method is intrinsically interpretable, and can be customized for different evaluation metrics. Our implementation is available at <jats:ext-link xmlns:xlink="http://www.w3.org/1
AU  - Zabeti,H
AU  - Dexter,N
AU  - Safari,AH
AU  - Sedaghat,N
AU  - Libbrecht,M
AU  - Chindelevitch,L
DO  - 10.1101/2020.05.31.115741
PY  - 2020///
TI  - INGOT-DR: an interpretable classifier for predicting drug resistance in M. tuberculosis
UR  - http://dx.doi.org/10.1101/2020.05.31.115741
ER  -
Download