Interpretable artificial intelligence and exascale molecular dynamics simulations to reveal kinetics: Applications to Alzheimer's disease

• Published in: Current opinion in structural biology Vol. 72; no. na; pp. 103 - 113
• Main Authors: Martin, William, Sheynkman, Gloria, Lightstone, Felice C., Nussinov, Ruth, Cheng, Feixiong
• Format: Journal Article
• Language: English
• Published: England Elsevier Ltd 01.02.2022; Elsevier
• Subjects: Alzheimer Disease - metabolism; Artificial Intelligence; BASIC BIOLOGICAL SCIENCES; Biological and medical sciences; Humans; Kinetics; Machine Learning; Molecular Dynamics Simulation
• ISSN: 0959-440X; 1879-033X; 1879-033X
• DOI: 10.1016/j.sbi.2021.09.001

The rapid increase in computing power, especially with the integration of graphics processing units, has dramatically increased the capabilities of molecular dynamics simulations. To date, these capabilities extend from running very long simulations (tens to hundreds of microseconds) to thousands of short simulations. However, the expansive data generated in these simulations must be made interpretable not only by the investigator who performs them but also by others as well. Here, we demonstrate how integrating learning techniques, such as artificial intelligence, machine learning, and neural networks, into analysis pipelines can reveal the kinetics of Alzheimer's disease (AD) protein aggregation. We review select AD targets, describe current simulation methods, and introduce learning concepts and their application in AD, highlighting limitations and potential solutions. •Expanding computing power allows for longer simulation times.•Machine learning techniques make these long simulations interpretable.•Kinetic properties of Alzheimer's disease proteins could lead to therapeutics.•Unsupervised learning has been used on multiple targets in Alzheimer's disease.