Evolutionary Medicine

Past selection has shaped present-day disease risk. In Souilmi et al., Current Biology (2021) we identified an ancient viral epidemic involving host coronavirus-interacting genes more than 20,000 years ago in East Asia, surfacing candidate genes relevant to modern coronaviruses; the paper remains the most publicly engaged piece in my catalogue (peak Altmetric 2,519, cited in the Finnish Prime Minister's Office COVID-19 research review). With Tobler and colleagues we extended this lens to the out-of-Africa expansion: genetic selection and climatic factors jointly shaped human dispersal (PNAS, 2023). I also led the demonstration that admixture has obscured signals of historical hard sweeps in humans (Nature Ecology & Evolution, 2022) — a methodological correction with implications for any selection scan that ignores admixture history.

Functional and clinical-translational work brings these signals into disease biology directly. With Sinniah and colleagues we showed in Nucleic Acids Research (2025) that conserved facultative heterochromatin marks regulatory sequences underpinning cell identity and disease — bridging population-scale evolutionary signal with mechanistic disease genomics. With Rogers and colleagues we characterised the impact of TNFAIP3 variation on NF-κB activation in acute kidney injury (Kidney International, 2023). And in a clinical-paleogenomics case study, the team integrated morphology, osteology, and ancient genomics to diagnose a 1,000-year-old case of Klinefelter's syndrome (The Lancet, 2022).

My current focus in this theme is Indigenous genomics, conducted under Indigenous governance and leadership. At BODL, I contribute genomics and bioinformatics expertise to community-led research using the largest cohort of Indigenous genomes in Australia, characterising the genetic determinants of Type 2 Diabetes in Aboriginal South Australians. The programme is supported by the MRFF Genomics Health Futures Mission grant "Pathways to benefit for Indigenous Australians in Genomic Medicine" (MRF2016124, A$4.99M), complemented by an MRFF-funded long-read nanopore sequencing platform for Indigenous genomics (MRF2016008, A$0.99M), and by the Australian Academy of Science Theo Murphy Initiative "Indigenous Genomics and Responsible Research" MOOC. Earlier collaborative work in this space includes Equitable Expanded Carrier Screening Needs Indigenous Clinical and Population Genomic Data (American Journal of Human Genetics, 2020) and the development of a draft Australian reference pangenome with the National Centre for Indigenous Genomics. Across all of it, the principle is the same: genomic methods are tools placed in service of Indigenous-led research priorities and community benefit.

Dingo Genomics & Conservation

Dingoes are an iconic, ecologically and culturally important free-living canid, and a focal organism for my conservation genomics work. In Souilmi et al., PNAS (2024), our team used ancient genomes to reveal over two thousand years of dingo population structure, providing a pre-Colonial baseline against which present-day admixture with domestic dogs can be quantified — directly informing dingo management and policy. The paper drew a fresh wave of tier-1 international coverage three years after the data were collected, and the 2026 commentary on gene-edited "super-quolls" (The Age, SMH, Brisbane Times, WA Today) picked up the same conservation-genomics thread. Earlier work in Zootaxa (2019) contributed to the long-running debate on dingo taxonomy.

On the methods side, our diploid dual canine assemblies (NAR Genomics and Bioinformatics, 2026) revealed extensive allelic heterogeneity and a telocentric chromosomal architecture — improving the genomic resources that underpin both wild- and companion-canid genetics. I also advise the Centre for Invasive Species Solutions and the South Australian Department of Primary Industries and Regions (PIRSA) on applied conservation and invasive-species genomics.

Bioinformatics, Methods & Cloud Infrastructure

Reproducibility and infrastructure underpin every theme above. I have worked with cloud-based scientific computing since 2012, when production-scale genomics on commercial cloud was still fringe. At Harvard Medical School I helped build the COSMOS distributed-pipeline framework and its bioinformatics pipelines (GenomeKey, PV-Key, and the multi-cloud variant MC-GenomeKey); these were later deployed at Invitae in clinical production, processing hundreds of thousands of patient samples — an early concrete demonstration that cloud-deployed genomic workflows could meet diagnostic-grade reliability.

At Adelaide I have played a key role in institutional adoption of AWS through the RONIN research-cloud platform, and have contributed open methods that the field now uses: the Genozip universal genomic-data compressor (Bioinformatics, 2021); systematic benchmarks of ancient-DNA read mapping (Briefings in Bioinformatics, 2021); the taxonomicfiltering Kraken2-based pre-mapping workflow that cuts ancient-DNA mapping runtime by up to ~94% (Briefings in Bioinformatics, 2024); and consensus recommendations for reproducible paleogenomic data analyses (American Journal of Human Genetics, 2025). My PhD student Shyamsundar Ravishankar won the AGTA 2024 best oral student presentation award for cloud-backed ancient-DNA workflows built on this stack.