Modeling Plasmodium falciparum diagnostic test sensitivity using machine learning with histidine-rich protein 2 variants

Search for this publication on Google Scholar
Ford, C., Janies, D., Alemayehu, G., Lo, E., Blackburn, K., Lopez, K., … Golassa, L. (2021). Modeling Plasmodium falciparum diagnostic test sensitivity using machine learning with histidine-rich protein 2 variants. Frontiers In Tropical Diseases. https://doi.org/doi:10.3389/fitd.2021.707313
Download PDF

Description

Malaria, predominantly caused by Plasmodium falciparum, poses one of largest and most durable health threats in the world. Previously, simplistic regression-based models have been created to characterize malaria rapid diagnostic test performance, though these models often only include a couple genetic factors. Specifically, the Baker et al., 2005 model uses two types of particular repeats in histidine-rich protein 2 (PfHRP2) to describe a P. falciparum infection, though the efficacy of this model has waned over recent years due to genetic mutations in the parasite. In this work, we use a dataset of 100 P. falciparum PfHRP2 genetic sequences collected in Ethiopia and derived a larger set of motif repeat matches for use in generating a series of diagnostic machine learning models. Here we show that the usage of additional and different motif repeats in more sophisticated machine learning methods proves effective in characterizing PfHRP2 diversity. Furthermore, we use machine learning model explainability methods to highlight which of the repeat types are most important with regards to rapid diagnostic test sensitivity, thereby showcasing a novel methodology for identifying potential targets for future versions of rapid diagnostic tests.

Details
Title

Modeling Plasmodium falciparum diagnostic test sensitivity using machine learning with histidine-rich protein 2 variants

Author(s)

Ford, Colby
Janies, Daniel
Alemayehu, Gezahegn
Lo, Eugenia
Blackburn, Kayla
Lopez, Karen
Dieng, Cheikh Cambel
Golassa, Lemu

UNC Charlotte Affiliate Type

Faculty

Extent

1 online resource

Date Published

2021

Material Type

Journal article

Genre

articles

Resource Type

text

Abstract

Malaria, predominantly caused by Plasmodium falciparum, poses one of largest and most durable health threats in the world. Previously, simplistic regression-based models have been created to characterize malaria rapid diagnostic test performance, though these models often only include a couple genetic factors. Specifically, the Baker et al., 2005 model uses two types of particular repeats in histidine-rich protein 2 (PfHRP2) to describe a P. falciparum infection, though the efficacy of this model has waned over recent years due to genetic mutations in the parasite. In this work, we use a dataset of 100 P. falciparum PfHRP2 genetic sequences collected in Ethiopia and derived a larger set of motif repeat matches for use in generating a series of diagnostic machine learning models. Here we show that the usage of additional and different motif repeats in more sophisticated machine learning methods proves effective in characterizing PfHRP2 diversity. Furthermore, we use machine learning model explainability methods to highlight which of the repeat types are most important with regards to rapid diagnostic test sensitivity, thereby showcasing a novel methodology for identifying potential targets for future versions of rapid diagnostic tests.

Subjects--Topics

Malaria
Plasmodium falciparum

Is Part Of

Frontiers in tropical diseases

Volume

Vol. 2

Language

English

Published Version (Please cite this version)

doi:10.3389/fitd.2021.707313

Publisher's Version

https://www.frontiersin.org/articles/10.3389/fitd.2021.707313/full

Permalink

http://hdl.handle.net/20.500.13093/work:249

Rights

This item may be protected by copyright and other related rights. Atkins Library provides access to this item for educational and research purposes only; other uses require the permission of the copyright holder.