Network models for psychopathology (e.g. comparing 'healthy' and depressed individuals' networks), and early warning signals for depression based on changes in autocorrelation. Both are a bit more complicated than simple differences in cross-sectional correlations though.