Table 1.
Statistical analysis methods and results for validation data
| Validation | Method | Model | Model output | Other statistical output | Number of samples (4 fields of view each quantified) | ||||
|---|---|---|---|---|---|---|---|---|---|
| SPARC + RBFOX1- OLIG2 + (IF) | Linear Model | Cube root (percentage of SPARC + RBFOX1-OLIG2 +) ~ tissue region + age*sex | Predictor | Est | SE | t-value | p-value | NA |
BA4: 6 CB: 4 CSC: 6 |
| (Intercept) | 1.2 | 0.3 | 4.06 | < 0.001 | |||||
| Tissue: CB | − 0.17 | 0.36 | − 0.47 | 0.64 | |||||
| Tissue: CSC | 0.74 | 0.32 | 2.34 | 0.023 | |||||
| Age Group: Young | − 0.43 | 0.46 | − 0.931 | 0.356 | |||||
| Sex: Male | − 0.02 | 0.39 | − 0.07 | 0.94 | |||||
| Young: Male | 0.35 | 0.61 | 0.578 | 0.565 | |||||
| HCN2 + SPARC + OLIG2 + (IF) | Wilcoxon Rank Sum test | Proportion of HCN2 + SPARC + oligodendrocytes ~ tissue | NA | W = 18.5 | p-value = 0.025 |
BA4: 3 CSC: 2 |
|||
| PAX3 + PDGFRA + (BaseScope duplex) | Poisson Model | Count of double positive cells ~ total count of PDGFRA positive + tissue | (Intercept) | − 1.31 | 0.34 | − 3.82 | < 0.001 | NA |
5 BA4 5 CB 7 CSC |
| Total count PDGFRA | 0.05 | 0.02 | 1.97 | 0.048 | |||||
| Tissue: CB | 1.50 | 0.31 | 4.87 | < 0.001 | |||||
| Tissue: CSC | 1.65 | 0.30 | 5.56 | < 0.001 | |||||
| NELL1 + PDGFRA + (BaseScope duplex) | Linear Model | NELL1 + PDGFRA + ~ tissue region | (Intercept) | 0.67 | 0.16 | 4.3 | < 0.001 | NA |
3 BA4 3 CSC |
| Tissue: CSC | − 0.67 | 0.22 | − 2.97 | 0.007 | |||||
| EBF1 + PDGFRA + (RNAscope) | linear model | Proportion EBF1 + PDGFRA + ~ age group + tissue + mean cellular autofluorescence | (Intercept) | 0.43 | 0.2 | 4.09 | < 0.001 | NA |
CB old: 4 CB young: 4 CSC old: 5 CSC young: 5 |
| Age Group | 0.18 | 0.05 | 3.64 | < 0.001 | |||||
| Tissue (CSC) | − 0.32 | 0.05 | − 7.05 | < 0.0001 | |||||
| Cellular auto-fluorescence | 44.00 | 26.27 | 1.68 | 0.099 | |||||
IF Immuno-fluorescence, CB Cerebellum, CSC Cervical spinal cord, Est Estimate, SE Standard Error