Periodic Reporting for period 1 - PvRecur (Using Plasmodium vivax genetic data to estimate the cause of recurrent vivax malaria)
Período documentado: 2023-11-07 hasta 2025-11-06
Malaria is a major global health challenge. It is caused by Plasmodium parasites transmitted from human-to-human by female Anopheles mosquitoes. The parasite species Plasmodium vivax is one of the two most important causes of malaria and also the most difficult to eliminate because of its ability to relapse: cause recurrent blood-stage infection (recurrence) following the activation of dormant liver-stage parasites. Blood-stage treatment failure (recrudescence) and new infectious mosquito bites (reinfection) also cause recurrence. Distinguishing among these three causes of recurrence is important in both epidemiological and clinical studies, especially those that aim to estimate the clinical efficacy of antimalarial treatment. However, there are no definitive biomarkers of relapse, recrudescence and reinfection (the 3Rs).
Genetic data on P. vivax parasites sampled from two or more blood-stage infections in the same person can be used to infer the cause of recurrence. Previously, we developed a statistical model capable of computing posterior 3R probabilities using data on fewer than 10 microsatellite markers. This was a breakthrough, showing that statistical genetic inference of P. vivax recurrence is feasible, but the prototypic model has two major drawbacks. Firstly, it is encoded in study specific scripts that are not easily accessible. Secondly, it does not scale to hundreds of markers typical of amplicon sequencing (AmpSeq) data, which many malaria studies are now generating.
PvRecur aims to transform P. vivax recurrence state inference by providing a open-source tool that all P. vivax researchers (epidemiologists and clinical trial analysts) can use to compute 3R probabilities from P. vivax AmpSeq data (objective 1); and by demonstrating the utility of that tool through its application to AmpSeq datasets on samples collected in the Solomon Islands, Peru, and Ethiopia (objective 2). By clarifying the causes of recurrence, PvRecur will improve our epidemiological understanding of P. vivax and enable more accurate treatment efficacy estimation, guiding the optimization of new treatment regimens. Given that nearly 2.5 billion people live at risk of P. vivax infection, efforts to improve P. vivax control and elimination have a substantial potential impact on global health.
Genetic data on P. vivax parasites sampled from two or more blood-stage infections in the same person can be used to infer the cause of recurrence. Previously, we developed a statistical model capable of computing posterior 3R probabilities using data on fewer than 10 microsatellite markers. This was a breakthrough, showing that statistical genetic inference of P. vivax recurrence is feasible, but the prototypic model has two major drawbacks. Firstly, it is encoded in study specific scripts that are not easily accessible. Secondly, it does not scale to hundreds of markers typical of amplicon sequencing (AmpSeq) data, which many malaria studies are now generating.
PvRecur aims to transform P. vivax recurrence state inference by providing a open-source tool that all P. vivax researchers (epidemiologists and clinical trial analysts) can use to compute 3R probabilities from P. vivax AmpSeq data (objective 1); and by demonstrating the utility of that tool through its application to AmpSeq datasets on samples collected in the Solomon Islands, Peru, and Ethiopia (objective 2). By clarifying the causes of recurrence, PvRecur will improve our epidemiological understanding of P. vivax and enable more accurate treatment efficacy estimation, guiding the optimization of new treatment regimens. Given that nearly 2.5 billion people live at risk of P. vivax infection, efforts to improve P. vivax control and elimination have a substantial potential impact on global health.
A user-friendly R package called Pv3Rs was built and released to the Comprehensive R Archive Network (CRAN) with extensive documentation and a vignette. A dedicated website attached to the development version of Pv3Rs, which is freely available on GitHub, includes a suite of articles describing how Pv3Rs was evaluated. The results are summarised in an article published in the journal Bioinformatics. Within Pv3Rs, an algorithm that permits recurrence state inference using data on hundreds of markers has been implemented. The algorithm decomposes genetic observations (alleles) assigned to parasites (a large space over which the prototype sums) into a cartesian product over per-marker assignments. The runtime of 3R probability computation scales linearly with the number of markers.
Pv3Rs has been used to compute 3R probabilities for samples collected in the Solomon Islands, Peru and Ethiopia. The analyses feature in two separate manuscripts currently under peer-review. The Ethiopian data set is coupled with mosquito feeding experiments, allowing for an investigation into the transmissibility of relapse. In the Solomon Islands (moderate transmission), approximately 40% of recurrences with Pv3Rs-computed probabilities are estimated to be reinfections. In Peru (moderate-to-high transmission), approximately 60% of recurrences with Pv3Rs-computed probabilities are estimated to be reinfection. In the Ethiopian study, only symptomatic recurrences were genotyped. Around 50% of symptomatic recurrences were classified as relapse and reinfection using Pv3Rs-calculated probabilities. Parasite and gametocyte densities were not notably different among the classifications and there was no evidence for a difference in the proportion of mosquitoes that became infected after adjusting for gametocyte density, age and fever.
Pv3Rs has been used to compute 3R probabilities for samples collected in the Solomon Islands, Peru and Ethiopia. The analyses feature in two separate manuscripts currently under peer-review. The Ethiopian data set is coupled with mosquito feeding experiments, allowing for an investigation into the transmissibility of relapse. In the Solomon Islands (moderate transmission), approximately 40% of recurrences with Pv3Rs-computed probabilities are estimated to be reinfections. In Peru (moderate-to-high transmission), approximately 60% of recurrences with Pv3Rs-computed probabilities are estimated to be reinfection. In the Ethiopian study, only symptomatic recurrences were genotyped. Around 50% of symptomatic recurrences were classified as relapse and reinfection using Pv3Rs-calculated probabilities. Parasite and gametocyte densities were not notably different among the classifications and there was no evidence for a difference in the proportion of mosquitoes that became infected after adjusting for gametocyte density, age and fever.
One of the aims of the MSCA-PF is to advance the career of the Fellow. I have recently secured a permanent research position at my host institution. This is partly a result of the MSCA fellowship, which I believe advanced my career in several ways. Even before receiving the fellowship, preparing its application was a valuable experience that I believe strengthened my application to the permanent research post. Receiving the fellowship reinforced my intellectual independence and established my financial autonomy; for example, the fellowship has supported my participation in multiple international conferences, building my professional network. The career development plan ensured my career objectives were SMART (specific, measurable, achievable, relevant, and time-bound) and that my expectations were clearly defined. Importantly, the fellowship also provided a secure two-year period during which I have had the intellectual freedom to develop my project proposal for the permanent position.