Fig. 4: Dot plot summarizing linear regression analyses between different types of latent variables and quantitative resilience trait (QRT). | npj Dementia

Fig. 4: Dot plot summarizing linear regression analyses between different types of latent variables and quantitative resilience trait (QRT).

From: Uncovering hidden factors of cognitive resilience in Alzheimer’s disease using a conditional-Gaussian mixture variational autoencoder

Fig. 4

For each feature—PC, t-SNE, UMAP, and latent variable (LV) from our C-GMVAE model, linear regression was performed with QRT as the response variable. R² values and adjusted p-values were computed separately for each subclass, with multiple testing corrected using the Benjamini–Hochberg method. Dot color represents the R² value (ranging from 0 to 1), and dot size reflects statistical significance. This visualization highlights both the strength and reliability of each association. Significance levels: ns not significant (adjusted p ≥ 0.05); *p < 0.05; **p < 0.01; ***p < 0.001; ****p < 0.0001.

Back to article page