Prediction and verification of the AD-FTLD common pathomechanism based on dynamic molecular network analysis

摘要

Step 1: significantly changed phosphorylation sites were selected from comprehensive phosphoproteome analysis with total cerebral cortex tissues of four AD and four FTLD mouse models at 1, 3, and 6 months of age. Step 2: pathological protein network of each mouse model at each time point was generated based on PPI database as described in the methods. Step 3: protein nodes changed in more than two models and possessing a high score (>3 SD) of betweenness were designated as core nodes. Step 4: edges to connect closely related nodes were selected by correlation analysis of three 3D vectors in which ratios of disease/control values at 1, 3, 6 months were used as x, y, z positions. We defined core edge when two 3D vectors of two nodes at the end had a high correlation (correlation value ≥ 0.9). Step 5: core network was made based on core nodes and core edges. Step 6: core networks of AD and FTLD were compared to generate AD core network, FTLD core network, and AD-FTLD common core network. Small circles were selected only by p values (p < 0.05 in Welch’s test) while large circles were selected also by q values (q < 0.05 in Welch’s test with post hoc BH procedure). AD-specific nodes, FTLD-specific nodes, and AD-FTLD common nodes are indicated with red, blue, and orange colors.