Connectome Gradient‐Based Subtyping of Major Depressive Disorder Reveals Distinct Neurobiological and Transcriptomic Signatures

Abstract

Background/Objective Major depressive disorder (MDD) exhibits significant heterogeneity, and identifying these distinct biological subtypes aids clinical intervention. In this study, we employed functional gradient and Hydra clustering methods to distinguish subtypes of depression. Method Imaging data were derived from a large‐scale, multicenter MDD imaging database, comprising 1067 patients with MDD and 907 healthy controls (HCs). Based on the principal functional gradient, the Hydra method was employed to distinguish subtypes of MDD. We employed the Mann–Whitney U‐test to compare functional gradient abnormalities across different subtypes of MDD while simultaneously conducting imaging transcriptomics analysis. Results Using principal functional gradients, we identified two subtypes of MDD. Compared with HCs, subtype 1 exhibited abnormal functional gradient values in the dorsal attention network (DAN), ventral attention network (VAN), limbic network (LIB), frontoparietal network (FTP), and default mode network (DMN); subtype 2 exhibited abnormal functional gradient values in the visual network (VIS), somatosensory network (SMT), DAN, VAN, and DMN. Transcriptomics revealed that subtype 1 functional gradient abnormalities were significantly correlated with theory mind, 5‐hydroxytryptamine (5‐HT) 1B; subtype 2 functional gradient abnormalities were significantly correlated with motor function, dopamine D2 receptor, and norepinephrine transporter (NET). Conclusions Our research deepens the understanding of the biological diversity of MDD, providing a biologically grounded framework for stratifying patients, which may inform the future design of hypothesis‐driven clinical trials targeting specific cognitive or sensorimotor circuits.