Rozier, A. | bioRxiv Science

bioRxiv2025

Ovarian development is driven by early spatiotemporal priming of the coelomic epithelium

Ovarian organogenesis relies on the coordinated specification of supporting and steroidogenic lineages from multipotent coelomic epithelium (CE) progenitors. However, it remains unclear how ovarian cellular diversity arises and whether CE progenitors are fate-biased before or after ingression into the gonad. We show that CE cells in fetal mouse and human ovaries are transcriptionally heterogeneous and spatially organized into subdomains already primed toward supporting or steroidogenic identities. CE priming is dynamic, influenced by proximity to the mesonephros, with a transient coexistence of both progenitor types before resolution toward a predominantly supporting-biased CE. In mice, delamination of primed CE cells seeds intragonadal niches that generate pre-granulosa and steroidogenic progenitors. We further demonstrate that fetal steroidogenic progenitors give rise to adult stromal and theca cells, and that granulosa cells have a dual origin from CE-derived and supporting-like cells. Together, these findings reveal a conserved, spatially encoded program of ovarian lineage specification.

developmental biology↗

bioRxiv2020

DeepCOMBI: Explainable artificial intelligence for the analysis and discovery in genome-wide association studies

Deep learning algorithms have revolutionized data science in many fields by greatly improving prediction performances in comparison to conventional approaches. Recently, explainable artificial intelligence (XAI) has emerged as a novel area of research that goes beyond pure prediction improvement. Knowledge embodied in deep learning methodologies is extracted by interpreting their results. We investigate such explanations to explore the genetic architectures of phenotypes in genome-wide association studies. Instead of testing each position in the genome individually, the novel three-step algorithm, called DeepCOMBI, first trains a neural network for the classification of subjects into their respective phenotypes. Second, it explains the classifiers decisions by applying layerwise relevance propagation as one example from the pool of XAI techniques. The resulting importance scores are eventually used to determine a subset of most relevant locations for multiple hypothesis testing in the third step. The performance of DeepCOMBI in terms of power and precision is investigated on generated datasets and a 2007 WTCCC study. Verification of the latter is achieved by validating all findings with independent studies published up until 2020. DeepCOMBI is shown to outperform ordinary raw p-value thresholding as well as other baseline methods. Moreover, two novel disease associations (rs10889923 for hypertension and rs4769283 for type 1 diabetes) were identified.

bioinformatics↗