Great Apes Chimp

Report card for Correlation and Tree Mapping on Chimp neocortex (Jorstad et al. 2023)

Overview

A taxonomy was initially built using the chimp neocortex single nucleus dataset. In building the taxonomy, 1000 binary marker genes were selected based on their gene expression from the single-cell transcriptome. Subsequently, the dataset was mapped to itself, termed self-projection, for evaluating the ideal performances of correlation and tree mapping algorithms.

Quantitative analysis

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

Annotaion F1-score
Cluster Correlation Mapping 0.890
Cluster Tree Mapping 0.868

Correlation Mapping

  1. Label-wise F1-score

  2. Confidence values for correctly and incorrectly assigned labels

  3. Confusion matrix (row-normalized)

Tree Mapping

  1. Label-wise F1-score

  2. Confidence values for correctly and incorrectly assigned labels

  3. Confusion matrix (row-normalized)