image representing the current explore topic

Marmoset Basal Ganglia

Report card for Correlation and Hierarchical Mappings on the Marmoset Basal Ganglia (et al. 2025)

Overview

The consensus basal ganglia taxonomy of Johansen, Fu et al 2025 integrates HMBA single-nucleus RNA sequencing (snRNA-seq) data from human, macaque, marmoset, and previously published mouse basal ganglia, with the goal of generating a consensus cell type taxonomy that can be widely adopted by the scientific community. By focusing on conserved marker genes and shared molecular profiles to supplement established names from the broader community, we have developed a standardized naming system that captures the evolutionary relationships and functional distinctions among basal ganglia cell types. The HMBA consensus basal ganglia taxonomy is designed to streamline communication, foster collaboration, and facilitate the development of novel research tools targeting specific cell types across multiple species.
 
Subsequently, the dataset was mapped to itself, termed self-projection, for evaluating the ideal performance of correlation and hierarchical mapping algorithms.
 

Quantitative analysis

The analysis evaluates the predictions of correlation and hierarchical mappings in determining cluster labels in a self-projection evaluation.

Annotaion Correlation
 F1-score
Hierarchical
  F1-score
Neighborhood 0.991 0.992
Class 0.989 0.991
Subclass 0.974 0.979
Group 0.942 0.947
Cluster 0.695 0.666

Correlation Mapping

    1.  Label-wise F1-score
    2.  Confidence values for correctly and incorrectly assigned labels
    3.  Confusion matrix (row-normalized)

 

 

 

 

Hierarchical Mapping

    1.  Label-wise F1-score
    2.  Confidence values for correctly and incorrectly assigned labels
    3.  Confusion matrix (row-normalized)