Decoding the piglet’s gut microbiome

Later, at Aarhus University, he completed a PhD in Food Science, focusing on how biofunctional plant compounds could help prevent diseases in pigs and chickens, such as post-weaning diarrhea in piglets. Today, as a postdoc in the GHH group (Gut and Host Health), his work has shifted toward the microbiome—a hidden ecosystem that shapes animal health in profound ways.

“Nutrition has always been my main interest,” Kevin says. “Over time, I realized the microbiome is the key to understanding why certain interventions work—or don’t.”

His current project, part of the nutritional pillar in PIG PARADIGM, explores whether supplementary nutrition during the suckling period can prime the piglet microbiome to reduce the risk of post-weaning diarrhea. By providing milk-based supplements and liquid creep feed before weaning, the hypothesis is that piglets will develop a more resilient gut environment, better prepared for the dietary transition. Fecal samples collected over time allow Kevin to trace how these interventions influence microbial communities.

But understanding the microbiome is not simple. For years, researchers relied on 16S rRNA sequencing to answer the question “Who’s there?”—identifying bacteria at the genus level. Kevin explains: “That approach is useful, but it’s often limited. We can’t tell which species or strains are there, and we don’t know what they’re capable of doing.” To go deeper, scientists turn to shotgun metagenomics, which sequences all DNA in a sample, revealing both taxonomic composition and functional potential. “With shotgun data, you don’t just see an E. coli—you can sometimes see which strain of E. coli. That matters for pig health.”

The catch?

Turning millions of DNA fragments into knowledge is a puzzle. “We have all the pieces, but we don’t know what the picture looks like,” Kevin says. “Bioinformatics tools help us put that puzzle together.” By assembling overlapping fragments, researchers can reconstruct near-complete microbial genomes—known as metagenome-assembled genomes, or MAGs. These often include organisms never seen before, highlighting the gaps in current databases. Feeding these new genomes back into reference libraries creates a feedback loop: the more data researchers contribute, the better the tools become for everyone.

This is where collaboration enters the picture. Kevin is working to integrate his findings with the bioBakery ecosystem, a suite of pipelines developed by teams at Harvard University and the University of Trento. Among these tools, MetaPhlAn stands out for its ability to profile microbial communities from shotgun data with remarkable accuracy, often down to species or strain level. HUMAnN complements this by mapping genes to metabolic pathways, revealing what functions microbes might perform. “We’re going from raw letters to actionable insight, and bioBakery helps us do that step by step,” Kevin explains.

Looking ahead: Spring in Trento

Kevin’s connection to the University of Trento began when assistant professor Francesco Asnicar, formerly presented at a PIG PARADIGM monthly webinar. Francesco’s group is a global leader in microbiome bioinformatics, known for developing and maintaining MetaPhlAn and other bioBakery tools. The collaboration will start with Kevin sharing MAGs from pig samples, which Trento researchers will help characterize and integrate into MetaPhlAn’s database. “If we start collaborating with the people who make these tools, we can have better tools informed by animal data—not just human microbiomes,” Kevin says.

The partnership also includes a planned research stay in Trento this spring, where Kevin will spend about a month working directly with the team. “We won’t finish everything in a month, but we’ll kick-start the right things—and keep building from there,” he notes. The goal is to apply the updated tools to his pig datasets and refine workflows for animal microbiomes. It’s a short visit, but an important step toward building long-term collaboration and advancing Kevin’s own expertise in high-level bioinformatics.

Shotgun metagenomics is resource-intensive, requiring external sequencing services and high-performance computing platforms like GenomeDK. There are also scientific debates about synthetic genomes—organisms inferred from DNA but never cultured. Kevin sees these challenges as opportunities. “The best way to tackle these issues is creating more information, improving the tools, and collaborating openly. That’s how animal microbiomes become primary members of the community in global research,” he says.

According to Kevin, this work is about more than pipelines and algorithms: It’s about giving scientists the ability to ask better questions and get clearer answers—whether that means identifying which strain of E. coli is present in a piglet’s gut or understanding how early nutrition shapes microbial functions that influence health, he says.  “For me, the collaboration with Francesco Asnicar’s team is not just a technical upgrade; it’s a way to ensure that animal microbiomes are represented in the tools that drive microbiome science forward”.

Fact box:

• 16S rRNA amplicon sequencing: Taxonomy at genus level; cost‑effective; limited functional insight.
• Shotgun metagenomics: All DNA; species/strain resolution; functional potential; resource intensive.
• MetaPhlAn (bioBakery): Taxonomic profiling via clade‑specific markers; often species/strain level.
• HUMAnN (bioBakery): Functional profiling; maps genes to pathways and contributions.
• MAGs: Metagenome‑assembled genomes reconstructed from shotgun data; key for discovering novel taxa.
• GenomeDK: High‑performance computing platform used to run bioinformatics pipelines at scale.