The number and volume of Multiple Sclerosis (MS) lesions alone are insufficient indicators of MS severity [1]. Recently, neurocomputational models have gained attention for simulating neural activity and fine-tuning parameters to better match empirical data [2]. However, some dynamical models fail to efficiently produce negative correlations obtained in neuroimaging signals. In [3,4], the authors introduced a 'functional connectivity (FC) informed structural connectivity (SC){\textquoteright} approach using tools from statistical mechanics. Such hybrid resting-state structural connectomes (rsSC) allow to infer the balance of excitatory and inhibitory activity between brain regions while they have been successfully used in whole brain dynamics to produce high quality simulated synthetic blood-oxygen-level-dependent (BOLD) signals [5]. In this study, we consider 13 healthy controls (HC) females and 25 female MS patients (mean age ± std: 44±12 and 48±11) using diffusion-weighted (DW) and fMRI images from a 3T scanner. The resting-state fMRI data is preprocessed and registered to the standard MNI space. BOLD signals are then averaged within AAL-defined atlas regions, and inter-region correlations are computed, resulting in 90 by 90 functional connectivity matrices. Similarly, the DW images were also registered with the standard MNI space and then underwent probabilistic tractography. This latter pipeline results in a set of 90 by 90 SC matrices. Combining the structural and functional matrices, we obtain hybrid rsSC connectomes which will allow us to measure differences in the excitation-inhibition ratio [4] between brain regions for the HC and MS groups in our datasets respectively.