DeepAMR for predicting co-occurrent resistance of Mycobacterium tuberculosis

Abstract

Motivation Resistance co-occurrence within first-line anti-tuberculosis (TB) drugs is a common phenomenon. Existing methods based on genetic data analysis of Mycobacterium tuberculosis (MTB) have been able to predict resistance of MTB to individual drugs, but have not considered the resistance co-occurrence and cannot capture latent structure of genomic data that corresponds to lineages. Results We used a large cohort of TB patients from 16 countries across six continents where whole-genome sequences for each isolate and associated phenotype to anti-TB drugs were obtained using drug susceptibility testing recommended by the World Health Organization. We then proposed an end-to-end multi-task model with deep denoising auto-encoder (DeepAMR) for multiple drug classification and developed DeepAMR_cluster, a clustering variant based on DeepAMR, for learning clusters in latent space of the data. The results showed that DeepAMR outperformed baseline model and four machine learning models with mean AUROC from 94.4% to 98.7% for predicting resistance to four first-line drugs [i.e. isoniazid (INH), ethambutol (EMB), rifampicin (RIF), pyrazinamide (PZA)], multi-drug resistant TB (MDR-TB) and pan-susceptible TB (PANS-TB: MTB that is susceptible to all four first-line anti-TB drugs). In the case of INH, EMB, PZA and MDR-TB, DeepAMR achieved its best mean sensitivity of 94.3%, 91.5%, 87.3% and 96.3%, respectively. While in the case of RIF and PANS-TB, it generated 94.2% and 92.2% sensitivity, which were lower than baseline model by 0.7% and 1.9%, respectively. t-SNE visualization shows that DeepAMR_cluster captures lineage-related clusters in the latent space. Availability and implementation The details of source code are provided at http://www.robots.ox.ac.uk/∼davidc/code.php. Supplementary information Supplementary data are available at Bioinformatics online.

Geographic heterogeneity impacts drug resistance predictions in Mycobacterium tuberculosis

“…This curve can be used to determine the optimal sensitivity for a given specificity, or vice versa. Furthermore, one can use the Area Under this Curve (AUC) to evaluate and compare the performance of different models, as is commonly done in drug resistance studies [10, 20, 21].…”

Use of Artificial Intelligence and Machine Learning for Discovery of Drugs for Neglected Tropical Diseases

“…They are being adopted rapidly for disease diagnosis and analysis of data from x-ray, CT, and MRI imaging. There are an increasing number of papers now employing deep learning methods for diagnosis of NTDs (e.g., Gao and Qian, 2018 ; Ting et al, 2018 ; Khalighifar et al, 2019 ; Rajaraman et al, 2019 ; Yang et al, 2019 ; Fuhad et al, 2020 ) that are beyond the scope of this review. Although deep learning methods offer state-of-the-art performance in modelling the biological properties of drug-like data for next generation drugs ( Ferreira and Andricopulo, 2019 ; Lavecchia, 2019 ), they have not yet been widely adopted by the NTD research community.…”

Robust detection of point mutations involved in multidrug-resistant Mycobacterium tuberculosis in the presence of co-occurrent resistance markers

“…Traditionally, estimating TB drug resistance has been based on known and biologically established mutations [3,10]. However, as resistance phenotype prediction from genomic data is a binary classification problem with high-dimensional input-a standard task in statistical and machine learning (ML)-various such techniques have been applied to antibiotic resistance recently [19][20][21][22][23][24][25][26]. Employing ML in genomic phenotype prediction has two main advantages.…”

“…Employing ML in genomic phenotype prediction has two main advantages. First, several recent studies have shown that these techniques can at least compete with existing direct association methods based on mechanistic and evidential knowledge, which has been curated and scrutinised for decades [20][21][22][24][25][26]. Second, examining important feature sets of the trained models might hint at yet unexplored variants leading to novel discoveries or reveal latent multidimensional interactions (e.g.…”