Scalable construction of near-wild-type brain atlases enables variability-aware neuronal identification in C. elegans

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.

Authors

Institutions

Publication Details

Journal
bioRxiv (Cold Spring Harbor Laboratory)
Published
2026-11-16
DOI
https://doi.org/10.64898/2026.11.12.766186
Primary Topic
Cell Image Analysis Techniques
Type
preprint

Funders

Controls
|||
ALL TIME
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
OCT
NOV
preprint

Scalable construction of near-wild-type brain atlases enables variability-aware neuronal identification in C. elegans

Zhenning Yang, Jingting Liang, Sihoon Moon, Hang Lu et al.
2 links
bioRxiv (Cold Spring Harbor Laboratory)
Cell Image Analysis Techniques
preprint

Scalable construction of near-wild-type brain atlases enables variability-aware neuronal identification in C. elegans

Zhenning Yang, Jingting Liang, Sihoon Moon, Hang Lu, Minghai Ge, Hyun Jee Lee, Yun Zhang, Wenjing Yang, Zikai Yu, Lizeth Parra Arroyo, Shiv Kumar, Yesol Moon, W. Alexander Calhoun
preprint en

Abstract

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.

bioRxiv (Cold Spring Harbor Laboratory)
Georgia Institute of Technology (US), Harvard University (US)
Simons Foundation
Cell Image Analysis Techniques
AI Navigator

Ask Laika to Summarize, Analyze, and Connect papers live on the map.

Summarize Papers & Methodologies

Extract key findings, datasets, and comparative methods across publications.

Benchmark Rankings & Visual Analytics

Rank top research institutions, authors, funders, topics, and journals by Field-Weighted Citation Impact (FWCI) and paper volume with instant charts.

Connect Distant Disciplines

Bridge topological clusters on the map to find hidden collaborative intersections.

Laika access coming soon