Photo of busy road crossing with pedestrians

How WWARN data helped build a global picture

Analysis from the MalariaGEN team at the Liverpool school of Tropical Medicine brings together the largest global dataset to date on kelch13 mutations linked to artemisinin resistance, building a comprehensive picture of antimalarial resistance emergence and spread. 

 

This study combines genomic data from MalariaGEN with tens of thousands of additional samples curated from the literature and public repositories – including major contributions from WWARN’s Artemisinin Molecular Surveyor database.

The problem

Artemisinin-based combination therapies (ACTs) pair a fast-acting artemisinin derivative with a longer-acting partner drug, ensuring malaria parasites are cleared and reducing the risk of resistance emerging in parasites. “They’re the most important malarial treatment at the moment,” says Dr Andrew Balmer, the first author of the study. 

But history offers a warning: malaria parasites successfully evolved resistance to chloroquine and sulfadoxine, previously effective frontline antimalarial drugs. This resistance first emerged in Southeast Asia before spreading to Africa, with devastating consequences. 

Currently effective ACTs may be heading the same way, with partial resistance to artemisinin (associated with mutations in the kelch 13 gene) first appearing in Southeast Asia, but now being spotted in parts of East and Northeast Africa. 

This is deeply concerning: Africa bears the overwhelming majority of global malaria cases and deaths, and if ACTs were to fail at scale in the continent, the human and economic costs would be catastrophic.

“There's a lot of data on partial resistance to combination therapies,” says Dr Andrew Balmer. 

“But it's been scattered across different journals, different datasets and different platforms. We wanted to aggregate all of that into one resource to create a global overview.”

Without a unified picture, it is difficult for policymakers and researchers to understand whether isolated reports represent noise – or the early stages of a continent-wide shift.

photo of long exposure road lights

The solution

To address this, Balmer and colleagues assembled what is so far the largest aggregation of kelch13 data to ever be analysed.

 

The team combined:

 

-  Whole genome sequence data from MalariaGEN’s Pf7 release: over 20,000 samples 

 

-  Over 70,000 additional curated samples from the WWARN database

 

- Tens of thousands of further samples carefully extracted from published literature

 

The final result was a single, large-scale, standardised dataset capable of supporting population-level analysis across time and geography.

Results

The final result was a single, large-scale, standardised dataset capable of supporting population-level analysis across time and geography.

We really valued the WWARN dataset.

Dr Cristina Ariani, study senior author.

This data was invaluable for us.

 Dr Andrew Balmer, study first author. 

By bringing all of these datasets together, the researchers were able to visualise trends over time and compare regions directly, allowing the data to speak for itself, rather than relying heavily on modelling assumptions.

What the data revealed

The analysis confirmed a worrying pattern: kelch13 mutations in East and Northeast Africa are increasing at a rate comparable to the early stages of resistance emergence previously seen in Southeast Asia.

But there are important differences too: in Southeast Asia, one mutation quickly became dominant. In contrast, in East Africa, multiple mutations appear to be emerging independently, suggesting that a single origin mutation is not spreading. 

The reasons for these differences remain uncertain. Variations in transmission intensity, host immunity, drug pressure, ecological factors, and health system dynamics may all play roles.

Three possible future scenarios

Photo of road against clear blue sky and dry brush

Sustained increase in resistance – resistance spreads more widely, partner drugs begin to fail, and ACT efficacy falls sharply.

 

Containment through intervention – aggressive surveillance, new treatment strategies (potentially including triple ACTs), vaccines, and coordinated mitigation slow or even prevent widespread treatment failure.

Heterogeneous outcomes – resistance emerges in some regions but not others.

 

The first scenario would carry enormous health and economic costs, and the authors argue that the next five years are critical in deciding which of these scenarios plays out. 

Why data sharing matters now

One of the striking findings from this large-scale analysis is the huge drop in publicly available sampling data after 2020. In some places, there are substantial temporal gaps, which means that we don’t know how malaria resistance is evolving. 

“There is typically a lag between sampling and data becoming publicly available,” notes Dr Richard Pearson, who was instrumental in developing the MalariaGEN parasite data resources, and is a co-author on the study. “We need to shorten that timeline, while ensuring that the scientists who collect and generate the data receive appropriate credit.”

The researchers highlight several priorities for future surveillance:

  • Longitudinal sampling in consistent geographic sites
  • Greater integration of whole genome sequencing, not just targeted kelch13 assays
  • Identification of markers of partner drug resistance
  • Expanded sequencing and bioinformatics capacity within endemic countries
  • Faster, equitable data sharing

Malaria molecular surveillance can be a particularly good tool for wide monitoring, because of its scalability: once sequencing pipelines and bioinformatics workflows are established, they can crunch through a large number of samples from many regions, without proportional increases in cost or complexity. 

But this can only happen if samples are collected for genomic surveillance, and the resulting data shared publicly. 

Public access to harmonised, high-quality molecular surveillance data enables researchers to detect patterns earlier, compare regions directly, and test competing hypotheses. It can help transform isolated signals into information that policy makers can act on.

A resource built for reuse

This study demonstrates how curated, standardised datasets can unlock insights that individual studies alone cannot provide.

By combining MalariaGEN sequencing data with WWARN’s aggregated database and published literature, the MalariaGEN team created a global perspective on artemisinin resistance that would not have been possible from any single source.

“We're looking forward to making use of these resources again in the future,” says Dr Balmer.

As artemisinin resistance evolves, the ability to integrate genomic data rapidly and equitably across continents may determine whether history repeats itself – or whether the global malaria community stays one step ahead.

Note: The MalariaGEN team was based at the Wellcome Sanger Institute until November 2025, at which point it migrated to the Liverpool School of Tropical Medicine.