Parkinson's disease (PD) is being increasingly recognized for its complex clinical symptoms, including cognitive impairment, which are linked to dynamic changes in brain network configuration and genetic susceptibility. However, the time-varying patterns of network reconfiguration in patients with PD and their relationship with gene expression profiles and cognitive function remain unclear. To address this gap, we analyzed modular variability (MV) in time-varying brain networks-comparing group differences in MV between PD patients and healthy controls (HC), and correlating MV with clinical measures-while gene sets associated with PD-related network changes were identified using Allen Brain Atlas data and assessed for biological relevance. Our results showed that PD patients exhibited decreased MV in several brain regions but increased MV in the right precuneus compared to HC; at the subnetwork level, they had reduced MV in the sensorimotor network and increased MV in the frontoparietal network. MV changes in specific regions correlated with cognitive and anxiety scores, while spatial transcriptome analysis linked gene expression to network reconfiguration in PD; further, genes involved in PD network reorganization were associated with mitochondrial function and nervous system development, affecting specific cell types. In conclusion, our study identifies distinct dynamic functional network abnormalities in PD, highlights the right precuneus as a potential key hub integrating molecular, metabolic, and network abnormalities into a cascade potentially driving PD-related cognitive decline, and establishes potential links between macroscopic modular dynamics and microscopic transcriptomes-collectively advancing insight into PD's underlying biological mechanisms.Copyright © 2025 The Authors. Published by Elsevier Inc. All rights reserved.