34 - Applying Computational Methods for Simulating Quaternary States of Proteins.
Description/Abstract/Artist Statement
Proteins are composed of amino acids bonded together to create a polypeptide chain, which acts as an essential part of all biological systems. The protein folding problem has been investigated for over fifty years and has evolved into three separate problems. One of those is the protein structure prediction problem, which involves computational methodology. Understanding precise protein structures will lead to a better understanding of how proteins function and, conversely, how mutations can lead to disease states. Predicting protein structures can also accelerate drug discovery research and lead to important scientific and medical advances. AlphaFold is a groundbreaking artificial intelligence program recently developed by Google’s DeepMind team to predict protein structures. AlphaFold uses neural networks and deep learning-based algorithms to predict protein structures given an amino acid sequence as input. The first aim of this research study was to gain expertise in using AlphaFold to predict known protein structures using a test set we constructed. This laid the foundation for all subsequent studies. The second aim, which is underway, expands this computational approach to predict the three-dimensional structure of human alpha-synuclein, which is proposed to consist of a multimeric state. Knowing the native structure will also facilitate the characterization of the protein function, which is not well understood. This aim also involves extending the application and present capabilities of AlphaFold. The initial experimental results indicate that alpha-synuclein can form a tetramer, and further analysis reveals the presence of stabilizing hydrophobic interactions.
Faculty Advisor/Mentor
Lesley Greene
Faculty Advisor/Mentor Department
Chemistry & Biochemistry
College Affiliation
College of Sciences
Presentation Type
Poster
Disciplines
Biochemistry | Bioinformatics | Computer Sciences | Structural Biology
34 - Applying Computational Methods for Simulating Quaternary States of Proteins.
Proteins are composed of amino acids bonded together to create a polypeptide chain, which acts as an essential part of all biological systems. The protein folding problem has been investigated for over fifty years and has evolved into three separate problems. One of those is the protein structure prediction problem, which involves computational methodology. Understanding precise protein structures will lead to a better understanding of how proteins function and, conversely, how mutations can lead to disease states. Predicting protein structures can also accelerate drug discovery research and lead to important scientific and medical advances. AlphaFold is a groundbreaking artificial intelligence program recently developed by Google’s DeepMind team to predict protein structures. AlphaFold uses neural networks and deep learning-based algorithms to predict protein structures given an amino acid sequence as input. The first aim of this research study was to gain expertise in using AlphaFold to predict known protein structures using a test set we constructed. This laid the foundation for all subsequent studies. The second aim, which is underway, expands this computational approach to predict the three-dimensional structure of human alpha-synuclein, which is proposed to consist of a multimeric state. Knowing the native structure will also facilitate the characterization of the protein function, which is not well understood. This aim also involves extending the application and present capabilities of AlphaFold. The initial experimental results indicate that alpha-synuclein can form a tetramer, and further analysis reveals the presence of stabilizing hydrophobic interactions.