Deconvolution of multiple infections in Plasmodium falciparum from high throughput sequencing data

• DOI: 10.1093/bioinformatics/btx530
• Publication Date: 2017-08-21T11:10:32Z
• Authors: Sha Joe Zhu; Jacob Almagro-Garcia; Gil McVean
• Keywords: 0301 basic medicine; 0303 health sciences; Coinfection; Plasmodium falciparum; High-Throughput Nucleotide Sequencing; Sequence Analysis, DNA; Original Papers; 3. Good health; 03 medical and health sciences; Haplotypes; Humans; Malaria, Falciparum; Software

Abstract

Motivation The presence of multiple infecting strains of the malarial parasite Plasmodium falciparum affects key phenotypic traits, including drug resistance and risk of severe disease. Advances in protocols and sequencing technology have made it possible to obtain high-coverage genome-wide sequencing data from blood samples and blood spots taken in the field. However, analyzing and interpreting such data is challenging because of the high rate of multiple infections present.

Results We have developed a statistical method and implementation for deconvolving multiple genome sequences present in an individual with mixed infections. The software package DEploid uses haplotype structure within a reference panel of clonal isolates as a prior for haplotypes present in a given sample. It estimates the number of strains, their relative proportions and the haplotypes presented in a sample, allowing researchers to study multiple infection in malaria with an unprecedented level of detail.

Availability and implementation The open source implementation DEploid is freely available at https://github.com/mcveanlab/DEploid under the conditions of the GPLv3 license. An R version is available at https://github.com/mcveanlab/DEploid-r.

Supplementary information Supplementary data are available at Bioinformatics online.