/ News, Research

Better classification with machine learning

Summarizing graph

Machine learning enhances the classification of unfolding patterns in proteins. (Image: V. Doffini, Department of Chemistry, University of Basel)

Researchers from the SNI network have introduced a new machine learning technique specifically designed to enhance the analysis of protein unfolding using Atomic Force Microscopy (AFM) data. The team of Prof. Michael Nash (University of Basel and ETH Zurich) recently published the work in the scientific journal “Nano Letters”.

This method stands out because the applied program repeatedly analyzes the data, allowing for more precise and efficient data classification. «A significant aspect of this approach is its ability to detect protein unfolding pathways that were previously unnoticed, providing deeper insights into protein interactions», comments Vanni Doffini, SNI PhD student and first author of the study.

The application of this method to AFM data is particularly important, as AFM is a critical tool in biophysics for studying the mechanical properties of proteins at the molecular level. This advancement not only refines protein unfolding research but also broadens our understanding of biological systems, showcasing the synergy between computational technology and biophysical research.

Original publication

Vanni Doffini, Haipei Liu, Zhaowei Liu, and Michael A. Nash
Nano Lett. 2023, 23, 22, 10406–10413, https://doi.org/10.1021/acs.nanolett.3c03026

Research group Prof. Michael Nash