8. Mar. 2018

One of the awarded projects that will be financed through a fund supporting interdisciplinary cooperation at the Masaryk University (GAMU) belongs to Kostas Tripsianes from CEITEC MU. He and his partner received 200 000 EUR  for developing a new methodology by “Pushing the limits in automated NMR structure determination using a single 4D NOESY spectrum and machine learning methods”.

The GAMU interdisciplinary grant offers the framework to bridge recent NMR developments (Tripsianes group – CEITEC) with several years of expertise in applications of artificial intelligence (Brazdil group – Faculty of Informatics). The present synergy has the potential to meet the need for a fast and efficient method for NMR structural studies, but also, to expand transformative technologies of machine learning in biomolecular research. 

NMR structure determination relies on recording a network of distance restraints from multi-dimensional spectra. Due to the complicated nature of the data, NMR data analysis is a laborious task and the road to NMR structure is generally a long one. The NMR assignment problem, if anything else, is a pattern recognition problem. Today, machines trained via machine learning in some scenarios, such as image recognition, perform better than humans. These algorithms build models that generalize well given large amounts of data by discovering patterns and trends in the data. Therefore, machine learning offers a natural solution to the NMR assignment problem if provided a large training set of experimental data deposited by NMR users over the years.