A deep learning-based approach for identifying new acetylcholinesterase inhibitors from mushroom molecular database for Alzheimer's disease

Alzheimer's disease (AD) is the most common type of dementia that gradually impairs a person's thinking and memory abilities to carry out even the most basic tasks and daily activities. Cholinesterase inhibitors (ChEIs) have been proven to alleviate AD symptoms and might even reduce the disease course, in which acetylcholinesterase inhibitors (AChEIs) represent a prospective area of AD drug development. However, the medication could have adverse side effects and cannot entirely halt the progression of the disease, limiting its efficacy. As a result, numerous attempts have been undertaken to search for alternative medicines with various structures and inhibitory actions using several computer techniques. A potential field of research is naturally derived-AChEIs because of fewer side effects on humans. In this study, candidates for AChEIs are found using a ligand-based virtual screening based on deep learning algorithms from a database that contains bioactive compounds from mushrooms (bacmushbase.sci.ku.ac.th). The evaluation metric, so called AUC (area under the curve), is predicted to be 0.98, suggesting the model’s performance in identifying AChEI candidates. Following that, the top ten predicted compounds are further validated using docking and molecular dynamics (MD) simulations. Then the binding energies and modes of the AChE and the inhibitors are compared with that of the FDA-approved drug (donepezil). These methods can accelerate the development process by assisting with screening and directing the selection of new inhibitor candidates.