Last Updated: 17/02/2023
Artificial Intelligence-based drug resistance screening of malaria parasites using ‘Read Until’
Objectives
To validate ‘Read Until’ as an artificial intelligence (AI)-based method to conduct drug 8 resistance profiling of malaria parasites using genomic DNA from whole blood.
Molecular markers of artemisinin resistance are emerging in P. falciparum populations throughout Papua New Guinea (PNG). First line treatment for P. falciparum malaria in PNG is Artemisinin Combination Therapy (ACT). There are several SNPs with the k13 gene that have been associated with the resistant phenotype4, and some of these including the most common, C580Y have emerged in PNG in recent years. An important next step in understanding resistance would be to genotype the samples showing the resistant phenotype in ring stage survival assays and to measure genotype-phenotype associations confirming whether k13 alleles contribute to increased ring stage survival.
Read Until was introduced in 2016 allowing for predefined selection criteria for the Nanopore flow cell machinery to accept matching or reject non-matching sequencing DNA fragments. Thus, there is an enrichment of target versus non-target reads in the final sequencing. This approach should allow the sequencing to be simplified and to target specific regions including drug resistant targets without requiring molecular biology training, enrichment/amplification protocols and without a priori knowledge of possible resistance-associated SNPs. Moreover, the sequencing cost can be reduced which is optimal for a range of projects and for in-country applications. Read Until could also provide faster and more accurate information for surveillance or diagnostics of malaria throughout malaria endemic regions.
The project will be divided into two laboratory experiments:
1. Optimising Read Until protocol to sequence specific targets of archived PNG ICEMR samples: The PNG samples will be used initially to troubleshoot and optimise the Read Until protocol for malaria samples. Libraries will be sequenced in two ways using duplicates: (i) complete WGS using standard protocols8 and (ii) using Read Until of specific targets including kelch13, crt, dhfr, dhps, mdr1 and other relevant targets. Six samples will be sequenced with complete WGS and Read Until for protocol optimisation and direct comparison between the two sequencing methods. Comparative analyses of the drug resistance profiles of the two sequencing methods will then be applied to identify the quality of high coverage Read Until data and SNPs extracted from the whole genome with lower coverage. The drug resistance profiles will also be compared to the 12 drug resistance assay developed by Dulcie Lautu a PhD student in the Barry lab using multiplex PCR and amplicon sequencing. The complete WGS samples will undergo whole genome amplification (WGA) as per optimised WGA protocols8 before MinION library preparation. Selective WGA (sWGA) protocol will be used to remove human DNA contaminants and improve amplification of P. falciparum as per protocols optimised in the Barry lab8. Read Until sequencing will require the installation of the client ‘Read Until Api’ to the MinION (https://github.com/nanoporetech/read_until_api). MinKNOW for MinION 20.06 or later is required and can be installed using python3. MinKNOW enables the researcher to specify the genomic targets of interest including k13, crt, dhfr, dhps and mdr1. The MinION can then accept matching sequences to these targets and output the resulting sequencing and reject the sequences that do not match to the specified targets.
2. Read Until will be used to sequence PNG samples collected in 2016 in the East Sepik Province where a strong population bottleneck has been previously identified, and nearby to Wewak where C580Y parasites have been identified. In addition, we will sequence a subset of samples from recent longitudinal cohort from Madang, where C580Y has also been detected in Therapeutic Efficacy Studies. The PNG samples will be sequenced according to the optimised protocol from experiment 1 and previously optimised protocols from the Barry Lab8. ~15 to 20 PNG samples from East Sepik Province, ~15 to 20 samples from Wewak and ~10 to 15 samples from Madang will be chosen for Read Until sequencing. A total of ~50 samples will be sequenced in this small pilot study to identify variants associated with artemisinin resistance. The proportion of samples chosen from each geographical location will enable comparison between drug resistant variants emerging at different locations. Comparative data analyses are similar for both experiments: The comparative data analysis will consist of comparing the read depth and coverage across each SNP in the Read Until sequencing and full sequencing on the MinION. A Principal Component Analysis (PCA) will be generated to clearly illustrate the associations among the drug resistant genes.
Nov 2021 — Oct 2023
$15,000
