Skip to main content
Fig. 1 | Translational Neurodegeneration

Fig. 1

From: Pathophysiological subtypes of mild cognitive impairment due to Alzheimer’s disease identified by CSF proteomics

Fig. 1

CSF proteomics identifies pathophysiological subtypes of MCIAD. a Study workflow with CSF samples from 45 MCIAD and 23 MCIOther patients from the CCC cohort. Proteomic characterization and data analysis consisted of PLS-DA followed by MCIAD patient clustering analysis for subsequent subtype characterization through system biology analysis. Two clusters were identified and validated by submitting the proteomic data of 92 MCIAD and 102 MCIOther patients from the EMIF-AD cohort to the same analysis. b Multivariate analysis using PLS-DA (both cohorts) classified the two groups of MCI patients. c Variable importance in projection (VIP) scores for the top 15 most important proteins to discriminate MCIAD from MCIOther. d GO enrichment analysis performed for “Decreased proteins” and “Increased proteins” showed enrichment for proteolysis, response to stimulus, complement activation, coagulation and response to wounding in both cohorts, among others. Similar or related Biological Processes have the same color. e, f PLS-DA classifying the different clusters of MCIAD patients after a cluster analysis using nNMF performed for a two-cluster solution (e) and proteins that better discriminate the two clusters in each cohort (f). g GO analysis indicating two major subgroups of MCIAD patients with decreased levels of proteins: one related to biological processes such as cell adhesion, coagulation, immune system and complement activation (Cluster 1) and the other to neurodevelopmental processes (Cluster 2) on both cohorts. h Comparison of AD biomarkers between Clusters from CCC (upper panel) and EMIF-AD (lower panel) cohorts. Box plots show that patients from Cluster 2 had higher levels of CSF pTau (left), CSF total Tau (t-tau, center) and CSF Aβ42 (right) as compared to Cluster 1 (independent sample t-test: *P < 0.05, **P < 0.01, ***P < 0.001)

Back to article page