Fig. 7: Pseudo-temporal ordering of cognitive dysfunction, brain pathologies, and protein modules.
From: Inferring protein expression changes from mRNA in Alzheimer’s dementia using deep neural networks
Protein modules and their expression levels were computed using SpeakEasy. Pseudo-temporal changes of clinical phenotypes and module expression were smoothed using LOESS (local polynomial regression) curve fitting. Protein modules were grouped into early, mid, and late based on the period when they altered 25% of the overall variation, which was indicated as a thick line. Protein module annotations were based on enrichment of gene ontology and brain cell-specific genes for proteomic modules.
