PROTrEIN: a new European network to study the contribution of artificial intelligence to proteomics data analysis
The team “Proteomics and Mass Spectrometry of Biomolecules” led by Odile Schiltz at IPBS, which hosts the Toulouse Proteomics Infrastructure and is part of the Proteomics French Infrastructure (ProFI), will participate in an innovative doctoral training network (ETN for “Educational Training Network” of Marie Skłodowska-Curie actions), together with 8 European academic partners and 2 industrial partners. This PROTrEIN network, funded by the European Union under the Horizon 2020 program, aims at training a new generation of PhD students in computational proteomics, by facilitating their career development, knowledge sharing and acquisition of new skills. The 10 partners, which belong to 7 European countries (Germany, Austria, Belgium, Denmark, Spain, Finland and France), will host and train 15 PhD students. They will work together for the organization of workshops and conferences, but also for the realization of a joint scientific project.
The work carried out by this network should bring significant scientific outcomes. It represents an opportunity to explore the contribution of modern, and machine learning based, data analysis methods, to the interpretation of mass spectrometry datasets obtained after analysis of protein samples. Over the past decade, major technological advances enabled the analysis of proteomes in an almost exhaustive manner. Several thousands of proteins and post-translational modification sites can now be identified and quantified in complex cellular systems, especially in clinical samples. Recent instruments have reached unprecedented levels of sensitivity, robustness and throughput, allowing these approaches to provide essential information for understanding physiopathological mechanisms. However, these technological advances come with a significant increase of the volume and complexity of the data generated, which has reached a critical level making human interpretation of the results almost impossible. Computer scientists must imagine novel algorithms capable of exploiting the new nature of spectral data. The recent democratization of computer-based machine learning methods opens up new perspectives for improving the IT solutions developed to date, and these approaches constitute the main scientific axis of this new European network.
At IPBS, two PhD students will be recruited in 2021 to work on these methods. These new developments will contribute to improve the Proline tool and the mzDB file format, developed in the team by David Bouyssié, in close collaboration at the national level with members of ProFI.