In this paper, we propose a bioinformatics-based method, which introduces thermodynamic measures and topological characteristics aimed to identify potential drug targets for pharmaco-resistant epileptic patients. We apply the Gibbs homology analysis to the protein–protein interaction network characteristic of temporal lobe epilepsy. With the identification of key proteins involved in the disease, particularly a number of ribosomal proteins, an assessment of their inhibitors is the next logical step. The results of our work offer a direction for future development of prospective therapeutic solutions for epilepsy patients, especially those who are not responding to the current standard of care.