Downregulation of Keratin Proteins in the Neuroendocrine System of Amyotrophic Lateral Sclerosis with PT150 as a Dual Neuroprotectant

Amyotrophic Lateral Sclerosis (ALS) is a neurodegenerative disease characterized by motor neuron degeneration. Although ALS biomarkers exist, they lack specificity and early-onset. ALS also has a 40% misdiagnosis rate. This paper proposes the use of a Graph Neural Network (GNN) to diagnose ALS, new ALS-specific biomarkers, and a therapeutic based on this group. Gene expression and proteomic datasets were sourced. A kNN data representation was used to train a GNN for disease diagnosis. Achieving 98% accuracy between ALS and control, and 85% between ALS patients and ALS mimics (the specific diseases within this class were not disclosed), outperforming existing approaches. Integrated Gradients then quantified each genes importance in the models decision. KRTAP3-1 and KRTAP9-8 were substantial in the models classification process due to involvement in cytoskeletal organization. In parallel, statistical proteomi analysis revealed the significance of KRT6A, KRT6B, and KRT4 due to involvement in cytoskeletal organization, tissue development, and cellular differentiation, all related to ALS. These keratin proteins were expressed in the pituitary gland, suggesting that their downregulation contributes to hormonal dysregulation in ALS. Once Glucocorticoid Receptor (GR) was found to be the reason for keratin downregulation, molecular docking identified PT150 as a therapeutic, binding with and inhibiting GR and Cyclooxygenase 2, also reducing neuroinflammation in ALS.