Dr Jacob Almagro-Garcia, Translational Research Lead at MalariaGEN, has been awarded a new Phase 3 grant from the Gates Foundation, marking an exciting next step in MalariaGEN's efforts to turn large-scale genomic and computational research into practical tools for malaria vaccine development.
The grant tackles a formidable challenge. The proteome of Plasmodium falciparum can generate millions of possible peptide candidates for a vaccine, but finding peptides that look promising computationally is a complex process. A strong candidate may need to be presented across diverse human HLA types, remain conserved across parasite populations, be expressed at the right stage of the parasite lifecycle, have supporting experimental evidence, avoid problematic cross-reactivity and also be practical to test in the laboratory.
The new phase of the grant will bring these different layers of evidence together, moving from an enormous search space towards focused, evidence-based sets of candidates for experimental validation.
From millions of possibilities to better decisions
MalariaGEN has spent years developing the infrastructure, datasets and analytical approaches needed to make sense of malaria parasite diversity at a global scale. Phase 3 of the Gates-funded programme will continue to extend that expertise into this translational challenge: how do we use population-scale data to help design better malaria vaccines?
At the centre of the programme is the evolution of the existing Pf-PeptideFilter framework into a scalable, population-aware decision platform for vaccine antigen and peptide prioritisation. The aim is to bring together evidence spanning parasite population genomics, HLA diversity, transcriptomics, proteomics, immunology, functional annotation and other relevant biological data.
This matters because promising vaccine targets have to fulfil many different constraints. A peptide that looks excellent against one parasite strain, for example, may be highly variable in natural parasite populations. Another may be strongly predicted to bind one HLA molecule but offer poor coverage across human populations. Others may look compelling computationally but have little evidence that they are expressed or biologically accessible at the stage of infection a vaccine is intended to target.
Data science at global scale
Doing this across an entire parasite proteome, many parasite populations, and hundreds of human HLA alleles quickly turns into a large computational problem. The programme will combine large-scale data engineering with statistical and ML approaches designed specifically for vaccine prioritisation. This includes developing ensembles of peptide-MHC prediction methods, improving how different sources of evidence are calibrated and combined, and exploring new ML approaches where they can genuinely improve candidate selection.
The team will also work closely with other malaria vaccine researchers to support real vaccine-design programmes. The programme will provide shared infrastructure, standardised methods and tailored analytical support. Different vaccine strategies will naturally ask different questions, so the platform is being designed around constraints rather than around a fixed goal. Researchers will be able to prioritise candidates according to the biological and experimental requirements of their own programmes. In addition, experimental results from collaborators can be brought back into the system, creating a feedback loop between prediction, prioritisation, validation and improvement.
Dr Almagro-Garcia said:
"This phase is about moving from evidence generation to evidence-led decision-making. By working closely with partners, we can turn complex genomic, proteomic and immunological data into practical outputs that help prioritise vaccine candidates. To do this, we’ll deploy cutting-edge machine learning and statistical methods and incorporate feedback from validation studies to continuously improve the platform."
An open resource for the malaria community
Beyond its immediate outputs, the grant places strong emphasis on openness and reuse. Tools, data, and insights generated through the programme will be made available to the wider research community, supporting broader efforts to develop peptide-based vaccines.
The expected impact is therefore twofold. In the near term, collaborating teams will receive carefully prioritised candidate sets and the evidence needed to take them forward experimentally. Over time, those analyses will contribute to a growing shared resource that can be applied by other researchers tackling similar challenges.
Looking ahead
We are entering a period in which the problem in biology is increasingly not a lack of data but knowing what to do with it. For malaria vaccines, that means moving beyond generating ever-larger collections of candidates and predictions towards integrating evidence, uncertainty and experimental feedback into decisions.
By combining MalariaGEN’s population-scale genomic resources and computational expertise with advanced analytics and close partnerships with experimental teams, this new phase aims to shorten the path from millions of possible candidates to the handful worth testing, further establishing MalariaGEN as a platform for translational research.
