EMOD with Full Parasite Genetics: A modeling framework for evaluating parasite genetic metrics for operational malaria molecular surveillance
Discuss this preprint
Start a discussion What are Sciety discussions?Listed in
This article is not in any list yet, why not save it to one of your lists.Abstract
Malaria molecular surveillance (MMS) is becoming increasingly common in endemic settings and has been proposed as a tool for monitoring parasite transmission to inform programmatic decision-making. However, the conditions under which parasite genetic metrics provide interpretable signals for broader use cases, such as assessing intervention impacts and detecting importation, remain under-characterized. We present EMOD with Full Parasite Genetics (FPG), a simulation framework designed to explore how parasite genetic metrics arise from transmission, intervention, importation, and sampling processes at programmatically relevant timescales. Using seasonal scenarios across a range of transmission intensities, we demonstrate three principal findings. First, genetic metrics can detect insecticide-treated net intervention impacts at seasonal and yearly timescales, but the strength, timing, and form of the relationship between genetic and epidemiological measures vary by metric and sampling timing. Second, importation can break the expected relationship between parasite genetic diversity from local transmission intensity at very low incidence, allowing low-transmission settings with substantial importation to maintain elevated diversity metrics. Third, convenience sampling practices, including sample size, collection timing, and the clinical composition of sampled populations, introduce non-random biases in genetic metric estimation in a way that obscures the true transmission signal. Together, these findings show that parasite genetic metrics can support operational surveillance, but that their interpretation depends on transmission context, importation, metric choice, and sampling design. EMOD FPG provides a framework for evaluating these dependencies in future setting-specific analyses and for guiding the interpretation of parasite genetic data across sites and over time.
Author summary
Malaria control programs are increasingly interested in using parasite genetic data to understand where transmission is changing, whether interventions are working, and whether infections are being imported from elsewhere. However, genetic data can be difficult to interpret because the patterns observed in sampled parasites are shaped not only by transmission, but also by importation, seasonality, and how samples are collected.
We developed a simulation model that tracks malaria parasite genomes through human infections, mosquito transmission, recombination, and field-like sampling. We used this model to evaluate how parasite genetic metrics behave under realistic programmatic conditions. We found that genetic metrics can detect the effects of an insecticide-treated net intervention campaign, but that the timing and strength of the signal depend on the metric used. We also found that importation can make low-transmission settings appear genetically diverse, complicating interpretation of genetic metrics as sole indicators of transmission. Finally, we showed that sample size, collection timing, and whether samples come from symptomatic or asymptomatic infections can systematically bias genetic estimates.
These findings suggest that parasite genetic data can support malaria surveillance, but only when interpreted through careful consideration of transmission context, importation, metric choice, and sampling design.
