Related Experiment Video
Updated: Jun 24, 2025
07:10
In Vesiculo Synthesis of Peptide Membrane Precursors for Autonomous Vesicle Growth
Published on: June 28, 2019
5.7K
An active machine learning discovery platform for membrane-disrupting and pore-forming peptides.
Alexander van Teijlingen1, Daniel C Edwards2, Liao Hu2
11Pure and Applied Chemistry, University of Strathclyde, 295 Cathedral Street, Glasgow, G1 1XL, UK. tell.tuttle@strath.ac.uk.
Physical Chemistry Chemical Physics : PCCP
|June 14, 2024
Summary
Machine learning accelerates the discovery of short pore-forming peptides (PFPs). This approach identified 71 potential PFPs, with experimental validation confirming their membrane-disrupting abilities, advancing bionanotechnology and therapeutics.
Area of Science:
- Biotechnology and Nanotechnology
- Biophysics
- Computational Biology
Background:
- Pore-forming peptides (PFPs) are crucial for bionanotechnology, influencing cell viability and transport.
- Controlling membrane transport is vital for cellular homeostasis and therapeutic applications like drug delivery.
- Short PFPs are less understood, yet offer potential for novel applications.
Purpose of the Study:
- To accelerate the discovery of novel pore-forming peptides (PFPs) using machine learning.
- To identify ultrashort PFPs with membrane-disrupting capabilities.
- To experimentally validate predicted PFPs and understand residue characteristics for pore formation.
Main Methods:
- Utilized an active machine learning approach to screen a vast octapeptide sequence space (25.6 billion sequences).
- Identified 71 potential PFPs computationally.
- Experimentally tested 13 predicted sequences for membrane-disrupting and pore-forming abilities.
Main Results:
- All 13 experimentally tested sequences exhibited predicted membrane-disrupting activity.
- One identified peptide formed highly stable pores.
- The study identified more ultrashort (8-residue) PFPs than previously known.
- Key residue characteristics favoring pore formation were elucidated.
Conclusions:
- Machine learning significantly accelerates PFP discovery.
- Short peptides can effectively form pores in cell membranes.
- The findings provide a powerful methodology for discovering PFPs for diverse applications, including nanoreactors and therapeutics.
