Enabling PROTEIN STRUCTURE PREDICTION with Artificial Intelligence at Rutgers and Beyond
December 9, 2021 @ 1:00 pm
Thursday, December 9th 2021 • 1:00 PM EST Registration (required and limited): go.rutgers.edu/dlqx9nfb
This Institute for Quantitative Biomedicine Crash Course will present a broad overview of how Artificial Intelligence/Machine Learning (AI/ML) methods are being used for de novo protein structure prediction and provide hands-on experience with both AlphaFold2 and RoseTTAFold.
CASP14 revealed that AlphaFold2, developed by Google DeepMind, Inc., can predict threedimensional structures of small globular proteins with accuracies comparable to experimental methods. RoseTTAFold, developed at the University of Washington/Howard Hughes Medical Institute, approaches AlphaFold2 in terms of prediction accuracy while requiring fewer computational resources.
In this Crash Course, expert speakers will provide a solid foundation on the role of AI/ML in structural biology and showcase ongoing research efforts at Rutgers. During the hands-on tutorial, participants will learn how to utilize these new computational tools to compute structure models from amino acid sequences and download precomputed structure models from the AlphaFoldDB database. Local computing resources (Rutgers University Amarel Cluster) and access to Google Colab and the RoseTTAFold server will be made available during the hands-on session.
Balamurugan Desinghu, Ph.D. and Stephen K. Burley, M.D., D.Phil.
CO-Sponsors: RCSB Protein Data Bank, Institute for Quantitative Biomedicine, Rutgers Office of Advanced Research Computing (OARC), Office of Research