Answers are grounded in this paper and cite their sources.
Matching cells across individuals is essential for comparing cellular data but becomes challenging when cell positions vary between animals. Here we introduce a framework for constructing probabilistic neuronal atlases by integrating partially overlapping observations from sparsely labeled, near-wild-type strains to reconstruct cellular organization from incomplete observations. Using five strains imaged without compression in a microfluidic array, we built an atlas of 129 head neurons (55 classes) that captures positional variability and infers spatial relationships even between neurons never observed together. We find that variability is spatially structured with neurons in compact regions being more stereotyped, while neuronal relative positions are more stable than their absolute positions. Exploiting these statistical relationships, variability-weighted identification improved top-1 accuracy to 86.3%, approaching human inter-annotator agreement, with the correct identity among the top three candidates in 98.2% of cases. Relationships stitched across strains reproduced observations in a held-out strain, and the atlas supplied correct identities in males and in pan-neuronal calcium imaging in vast majority of cases. Our extensible framework enables community-built atlases and offers a strategy for cell identification when anatomical variability limits template-based registration.
No takes yet. Share an insight, caveat, or question.
Yu et al. (2026) studied this question.
Synapse has enriched 8 closely related papers on similar clinical questions. Consider them for comparative context: