Efficient Development of Orally-Bioavailable Cyclic-Peptide Drugs via Molecular Dynamics + Machine Learning
Cyclic peptides are a promising drug modality, combining beneficial traits from both small molecules and antibody or protein therapeutics. While many scientists think that cyclic peptides are key to providing orally-available therapeutics, there are only a few FDA-approved cyclic peptides due to the complexities of cyclic peptide drug discovery.
One of the main challenges in developing efficient cyclic peptides is due to their chameleonic properties as they tend to adopt multiple confirmations in a solution, which are difficult to characterize and predict.
Dr. Yu-Shan Lin’s team at Tufts University has combined molecular dynamic simulations to train machine learning models to efficiently predict cyclic peptide confirmation. This method is called StrEAMM (Structural Ensembles Achieved by Molecular Dynamics and Machine Learning). stream is the first technique capable of efficiently predicting complete structural ensembles of cyclic peptides without relying on additional molecular dynamics simulations, constituting a seven-order-of-magnitude improvement in speed while retaining the same accuracy as explicit-solvent simulations. Furthermore, the team is actively working to use structured data to train a predictive model that connects the confirmation to desirable drug properties like binding affinity and membrane permeability.
Distills experimental data to provide a predictive model (ML) and suggest next-generation designs
Enables rational design of cyclic peptides with desired structures and favorable drug properties
The prediction for each individual cyclic peptide can be made in less than 1 second of computation time, dramatically reducing cycle time
PUBLICATION: Structure prediction of cyclic peptides by molecular dynamics + machine learning (November 2021).
IP STATUS: Two provisional applications were filed (6/14/2021 & 10/14/2021)
TTO Home Page: http://tufts.technologypublisher.com
Name: Emma Anderson
Title: Licensing Associate