📢 Excited to share our latest work, now on bioRxiv! We've used deep learning to uncover functional archetypes in the adult human gut #microbiome, shedding light on interindividual variability microbiome stability, and disease associations: https://www.biorxiv.org/content/10.1101/2025.01.29.635381v1 #AI #microbiology 🧵 1/n
Comments
Gut microbiome composition varies widely across individuals. While #enterotypes and #enterosignatures offer compositional classifications, they don’t fully explain functional states. We need a better way to connect microbiome function with health and disease 2/n
We combined #deep-learning with archetypal analysis to define functional archetypes—extremal profiles that represent boundaries in microbiome functionality. Unlike taxonomy-based methods, this framework identifies core metabolic strategies. 3/n
1️⃣ Gut microbiomes fall into three functional archetypes with distinct metabolic pathways
2️⃣ Proximity to archetypes correlates with microbiome stability
3️⃣ Disease signals are confounded by archetypes
4️⃣ Targeting function could improve microbiome-based interventions 4/n
*Archetype 1: High carbohydrate metabolism → BCAA & microbial cell wall biosynthesis
*Archetype 2: Fatty acid metabolism + TCA cycle → highest metabolic flexibility = most stable & diverse
*Archetype 3: Amino acid + nitrogen metabolism → linked to inflammatory pathways
* Carb-heavy diets? → Likely Archetype 1
* Fatty-acid rich diets? → Archetype 2
* Protein-rich diets? → Archetype 3
6/n
7/n