Incorporating the missing time dimension into identity-by-descent-based malaria genomic surveillance

The National Institute of Environmental Health Sciences awarded $1.399 million to the University of Maryland, Baltimore on September 1, 2026, under the Medical Library Assistance program (CFDA 93.879) to incorporate the missing time dimension into identity-by-descent-based malaria genomic surveillance.

The project will develop improved analytical methods that integrate sampling time information into recombination-based metrics, specifically identity-by-descent (IBD) analysis, for malaria parasite population genetics. Current IBD-based genomic surveillance methods treat parasite populations as time-homogeneous and do not directly model sampling time, which creates a significant limitation: parasites sampled over a single decade may encompass genomes from up to 70 different generations, potentially confounding interpretations about transmission patterns and reducing statistical power to identify epidemiologically actionable signals. By incorporating temporal information—drawing on approaches used in ancient DNA studies for human population demography—the methods will improve accuracy and precision of demographic estimates, better characterize parasite migration and transmission networks, differentiate local from imported infections, and identify drug resistance markers and distinct parasite genetic subpopulations. These enhanced surveillance tools will support malaria control and elimination efforts.

Performance runs from September 1, 2026, through August 31, 2030, at the Baltimore campus location in Maryland.

Abstract

Intensive malaria control and elimination efforts have led to substantial reductions in the malaria burden globally over the last two decades; however, progress toward malaria elimination has stalled in recent years. The capacity to sequence malaria parasite genomes at epidemiological scales has heightened interest in utilizing genomic surveillance to complement traditional surveillance methods. Analyzing parasite genomic data using recombination-based metrics, such as identity-by-descent (IBD), provides key information about the demographic history of parasites on a finer scale and in more recent evolutionary timeframes. IBD-based methods are employed to understand reduced transmission intensity following interventions, identify drug resistance markers, detect distinct parasite genetic subpopulations and parasite migration, and describe transmission networks to differentiate local transmission from imported infections. However, existing methods often treat parasite populations as time-homogeneous and do not directly model sampling time, assuming that IBD segments of the genome are shared between parasites at the present time. Indeed, parasites sampled over only a decade may encompass genomes from up to 70 different generations, potentially confounding data interpretation related to time and likely reducing statistical power to identify more transient and epidemiologically actionable malaria transmission patterns. Incorporating sample time information has the potential to improve the accuracy and precision of demography estimates, as "ancient DNA" studies have informed estimates of human population demography. In addition, existing approaches to estimate parasite migration provide insight into malaria transmission patterns retrospectively. Such analyses reflect source/sink dynamics from the recent past but may be less informative about future patterns in parasite populations that evolve quickly. Therefore, models that can predict future patterns from existing data are critical for planning elimination strategies, especially for parasite populations under substantial evolutionary pressure due to malaria interventions. In the proposed study, we will develop a new IBD-based estimator of recent effective population size (𝑁𝑁𝑒𝑒), utilizing IBD segments from parasite genomes with heterogeneous sampling times. We will benchmark the new tool against existing alternative 𝑁𝑁𝑒𝑒 estimators to assess how incorporating sampling time information improves the accuracy of demography estimates (aim 1). We will also perform network analysis and link prediction using deep learning models to forecast future parasite population migration and the spread of drug resistance, using time-series data (aim 2). By incorporating sampling time into our analytical framework, we will be able to more precisely quantify the impact of malaria interventions on parasite population demography and identify transmission sources and sinks to inform targeted malaria control.