German, C. | Search bioRxiv

bioRxiv2026

Microbial Growth in an Enceladus Ocean Analog Medium Informed by Mineral Stability Modeling

Evidence from the Cassini mission confirmed that Saturns moon Enceladus hosts a subsurface alkaline ocean where rock-water reactions may generate redox disequilibria capable of supporting microbial metabolisms. To investigate potential microbial survival under simulated Enceladus ocean conditions, we used thermodynamic modeling to develop a salt formulation consistent with one possible Enceladus ocean composition and supplemented it with putative microbial energy sources to create a growth medium. The medium was inoculated with samples from diverse ocean world analog environments on Earth to determine which microorganisms could persist under Enceladus-like conditions. The microorganisms persisting in this geochemically bounded medium were heterotrophic, metabolically versatile bacteria with low carbon requirements. Genomic and physiological analyses further showed the presence of multiple stress-response pathways, sodium-based bioenergetic systems, osmoregulation strategies, and other adaptations consistent with survival in alkaline, low-nutrient settings. These results suggest that some stress-tolerant heterotrophic bacteria may serve as useful model organisms for life in Enceladus subsurface ocean. These findings demonstrate the value of geochemically modeled media as a framework for constraining habitability, identifying relevant biosignatures, and probing potential microbial survival strategies beyond Earth.

microbiology↗

bioRxiv2025

PRSformer: Disease Prediction from Million-Scale Individual Genotypes

Predicting disease risk from DNA presents an unprecedented emerging challenge as biobanks approach population scale sizes (N > 106 individuals) with ultra-high-dimensional features (L > 105 genotypes). Current methods, often linear and reliant on summary statistics, fail to capture complex genetic interactions and discard valuable individual-level information. We introduce PRSformer, a scalable deep learning architecture designed for end-to-end, multitask disease prediction directly from million-scale individual genotypes. PRSformer employs neighborhood attention, achieving linear O(L) complexity per layer, making Transformers tractable for genome-scale inputs. Crucially, PRSformer utilizes a stacking of these efficient attention layers, progressively increasing the effective receptive field to model local dependencies (e.g., within linkage disequilibrium blocks) before integrating information across wider genomic regions. This design, tailored for genomics, allows PRSformer to learn complex, potentially non-linear and long-range interactions directly from raw genotypes. We demonstrate PRSformers effectiveness using a unique large private cohort (N {approx} 5M) for predicting 18 autoimmune and inflammatory conditions using L {approx} 140k variants. PRSformer significantly outperforms highly optimized linear models trained on the same individual-level data and state-of-the-art summary-statistic-based methods (LDPred2) derived from the same cohort, quantifying the benefits of non-linear modeling and multitask learning at scale. Furthermore, experiments reveal that the advantage of non-linearity emerges primarily at large sample sizes (N > 1M), and that a multi-ancestry trained model improves generalization, establishing PRSformer as a new framework for deep learning in population-scale genomics.

genetics↗