Obsessive-Compulsive Disorder (OCD) entails intrusive thoughts and repetitive behaviors fueled by anxiety, influenced by cognitive, emotional, and behavioral factors. Medication is employed to address these dynamics. Recent theories explore a potential correlation between OCD symptoms and brain attractor network states, pointing to a connection with glutamatergic activity. In our investigation, conducted over three time points with 7-week intervals, we collected a comprehensive dataset comprising resting-state and task-based fMRI signals, DTI data, and metabolic details, particularly focusing on glutamate concentration. A group of 58 OCD patients and 62 healthy controls underwent memantine or placebo treatment for over 4 months. Utilizing a modified state space model integrated with a Recurrent Neural Network, we elucidated the evolving neurobiological landscape in OCD patients. The generative recurrent neural network captured dynamical features and attractor state changes. Results indicated that glutamate modulation induced alterations in behavioral patterns, cognitive parameters, and various structural and functional aspects of neuroimaging, including large-scale brain networks and neurodynamical variables like network attractors. Furthermore, the study emphasized that baseline connectivity correlated with the extent of memantine-induced changes across structural, metabolic, and functional connectivity, encompassing static, dynamic, and effective aspects. Correlations were observed between memantine-induced changes and spatial gene expression patterns of serotonin and glutamate receptors in a time and location-dependent manner. Simulated experiments demonstrated how attractor overstability or instability could impact network dynamics, providing insights into cognitive deficits, response inhibition, flexibility in plan shifting, and the manifestation of symptoms like obsessions and compulsions in OCD.