Towards a clinically-based common coordinate framework for the human gut cell atlas: the gut models

Area of Science:

• Computational biology
• Anatomical modeling
• Bioinformatics

Background:

• The Human Cell Atlas aims to spatially map single-cell transcriptomes.
• Understanding disease requires sophisticated spatial descriptive frameworks for pathological phenotypes.
• Current methods lack integrated spatial analysis for cellular data.

Purpose of the Study:

• To develop a conceptual coordinate model for the Gut Cell Atlas.
• To represent gut locations using a 1-dimensional (1D) linear model based on anatomical landmarks and distances.
• To enable integration and analysis of spatial data for the human gut.

Main Methods:

• Developed a Gut Linear Model (1D) based on the gut's centerline and standardized anatomical ontology terms.
• Incorporated landmarks (e.g., ileo-caecal valve) and distance measures for location semantics.
• Demonstrated mapping between 1D, 2D, and 3D models, including patient CT scans.

Main Results:

• Generated 1D, 2D, and 3D models of the human gut.
• Provided models as open-source Json and image files.
• Developed a demonstrator tool for exploring anatomical space and model mappings.

Conclusions:

• A 1D centerline model with landmarks accurately represents gut anatomy and function.
• This model facilitates interoperable translation between 1D, 2D, and 3D representations.
• Enables precise sample localization for comparative data analysis.