Abstract

The determination of lineages from strain-based molecular genotyping information is an important problem in tuberculosis. Mycobacterial interspersed repetitive unit-variable number tandem repeat (MIRU-VNTR) typing is a commonly used molecular genotyping approach that uses counts of the number of times pre-specified loci repeat in a strain. There are three main approaches for determining lineage based on MIRU-VNTR data - one based on a direct comparison to the strains in a curated database, and two others, on machine learning algorithms trained on a large collection of labeled data.All existing methods have limitations. The direct approach imposes an arbitrary threshold on how much a database strain can differ from a given one to be informative. On the other hand, the machine learning-based approaches require a substantial amount of labeled data. Notably, all three methods exhibit suboptimal classification accuracy without additional data.We explore several computational approaches to address these limitations. First, we show that eliminating the arbitrary threshold improves the performance of the direct approach. Second, we introduce RuleTB, an alternative direct method that proposes a concise set of rules for determining lineages. Lastly, we propose StackTB, a machine learning approach that requires only a fraction of the training data to outperform the accuracy of both existing machine learning methods.Our approaches demonstrate superior performance on a training dataset collected in New York City over 10 years, and the improvement in performance translates to a held-out testing set. We conclude that our methods provide opportunities for improving the determination of pathogenic lineages based on MIRU-VNTR data.

Highlights

  • The genetic diversity of the infectious pathogen Mycobacterium tuberculosis has played an important role in its adaptation to its diverse host species, including humans [1, 2, 3]

  • We conclude that our methods provide opportunities for improving the determination of pathogenic lineages based on MIRU-VNTR data

  • As we describe in the Supplementary Materials, Malioutov and Varshney [23] propose an approach based on linear programming (LP) to the NP-hard problem [24] of identifying the smallest set of complex rules of this form

Read more

Summary

Introduction

The genetic diversity of the infectious pathogen Mycobacterium tuberculosis has played an important role in its adaptation to its diverse host species, including humans [1, 2, 3]. Several different molecular genotyping methods have been used to assign lineages to M. tuberculosis, including restriction fragment length polymorphism (RFLP), spacer oligonucleotide typing (spoligotyping), large sequence polymorphisms (LSPs), single nucleotide polymorphisms (SNPs), and mycobacterial interspersed repetitive unit-variable number tandem repeats (MIRU-VNTR). The method consists in assigning to a strain of interest the lineage of the strain in the database that differs from it in the smallest number of loci, provided that this number does not exceed 4 out of 24 loci Another widely used method, called TB-Insight [21], uses a machine learning method called Conformal Bayesian Networks for the classification problem. We separate it into a training set, which we use to develop our methods, and a testing set, which we use to evaluate their performance

Dataset preparation
Removing the arbitrary threshold
Producing interpretable rules
Designing a machine learning method
Implementation
Model Training
Performance Analysis
Sensitivity Analysis
Performance of our methods on broad lineages
Performance of our methods on a refined classification
Comparison to existing methods
Discussion
Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call