- A comprehensive LFQ benchmark dataset on modern day acquisition strategies in proteomics : article - Puyvelde, Bart Van; Daled, Simon; Willems, Sander; Gethings, Lee A.; Bloomfield, Nic; Schiltz, Odile; Deforce, Dieter; Dhaenens, Maarten; Martens, Lennart; Tate, Stephen; Perez-Riverol, Yasset; Gabriels, Ralf; Gonzalez de Peredo, Anne; Chaoui, Karima; Mouton-Barbosa, Emmanuelle; Bouyssie, David; Boonen, Kurt; Hughes, Christopher J. - SpringerNature, 2022. - 12 pp..
- Interpretation of the DOME Recommendations for Machine Learning in Proteomics and Metabolomics : article - Palmblad, Magnus; Böcker, Sebastian; Degroeve, Sven; Kohlbacher, Oliver; Käll, Lukas; Wilhelm, Mathias; Noble, William Stafford - Washington : ACS Publications, 2022. - 4 pp. - (Journal of Proteome Research ; 2022).
- Sensitive and Specific Spectral Library Searching with CompOmics Spectral Library Searching Tool and Percolator : article - Shiferaw, Genet Abay; Gabriels, Ralf; Bouwmeester, Robbin; Bossche, Tim Van Den; Vandermarliere, Elien; Volders, Pieter-Jan; Martens, Lennart - Washington : ACS Publications, 2022. - 6 pp. - (Journal of Proteome Research ; 21. 2022, pp. 1365-1370).
The field of proteomics continues to evolve at a rapid pace, and as it does so, it becomes more and more of a data science. Driven by methodological changes that focus increasingly on applications outside of the traditional aim to catalogue protein presence or absence, the field is in constant need of advanced informatics approaches to make sense of the acquired data. The importance of these proteomics informatics solutions is now such, that without these computational tools, the field would no longer be able to function. As a result, the field of proteomics often finds itself limited by available algorithms and computational solutions, rather than by available instruments or samples.
Based on recent developments in the field, we have identified four highly interesting computational challenges (and thus opportunities) that have come to the foreground in proteomics: (i) algorithms for the interpretation of proteomics analyses of protein structure; (ii) computational support for the newly emerging field of single-cell proteomics analyses; (iii) the identification and localization of protein glycosylation using mass spectrometry; and (iv) the quickly emerging application of Machine Learning applications throughout the field of proteomics.
While the first three challenges are clearly matched to a specific application domain of proteomics (protein structure elucidation, single-cell analysis, and glycoprotein characterization), the fourth challenge, the application of Machine Learning algorithms, presents a broadly cross-cutting development in the field, which has already started to supply innovative solutions to long-standing problems in various areas of proteomics.
The primary goal of this interdisciplinary Dagstuhl Seminar, which brings together a mix of experts from different but relevant backgrounds around these four topics, is to exchange ideas that can lead to novel algorithms and tools, and that can create novel collaborations between various participants.
A strong focus throughout the seminar will therefore be on the free exchange of ideas between the various types of expertise that will be represented among the participants, to maximally benefit from obvious as well as less obvious synergies and to provide maximal opportunity for cross-fertilization of ideas. The structure of the seminar reflects this, and consists of a mix between thought-provoking plenary lectures, and dedicated breakout sessions that have smaller groups of participants focusing on specific core questions around the four topics.
The Dagstuhl Seminar 21271 "Computational Proteomics" discussed several important developments, challenges, and opportunities that are emerging in the field of computational proteomics. Three core topics were set out at the start, and these were discussed at length throughout the seminar.
These three topics were: (i) the fast evolving use of advanced machine learning approaches in proteomics; (ii) the challenges and opportunities offered by fast developing approaches for structural and top-down proteomics; and (iii) specific issues and computational complications in glycoproteomics.
The machine learning and glycoproteomics topics were each introduced by a dedicated lecture, which set out the current state-of-the-art and presented a tentative set of issues, challenges, or opportunities that could be explored during the seminar. The structural and top-down proteomics topic was introduced by two sequential lectures, one on structural proteomics, and one on top-down proteomics. In total, four introductory talks were thus presented at the start of the seminar. For each of the three main topics, daily Working Group sessions were organised, which took place in the morning and afternoon, with a daily late-night session scheduled each day to wrap up the day's outcomes. This structure was followed to allow maximum involvement by online participants across the various timezones in the hybrid format. The Machine Learning in Proteomics Working Group also spun out another Working Group session during the seminar, which discussed the creation of a machine learning (Kaggle-like) competition based on proteomics data.
Each of these breakout sessions was very actively attended, including by online attendees, and resulted in several interesting research ideas and potential new initiatives. The Machine Learning in Proteomics Working Group was the largest working group, and addressed a number of distinct topics during the seminar. Of particular note were the spin-out effort to establish two machine learning competitions based on proteomics data and challenges to engage the broader machine learning community, and the extensive discussions on the optimal way to represent mass spectrometry data for downstream machine learning.
The Glycoproteomics Working Group was very actively attended, and discussed an exciting set of topics. A first highlight among these topics was provided by the extensive and detailed discussions with the Machine Learning Working Group regarding the potential of, and road towards, the use of state-of-the-art machine learning approaches in glycoproteomics. A second highlight concerned the delineation of a set of high-impact opinion papers to describe the state-of-the-art of the field, and its goals, ambitions, and challenges.
The Structural and Top-Down Proteomics Working Group was very active in detailing the many challenges and opportunities in this fast-evolving field. One noteworthy challenge revolved around the detection, annotation, and biological interpretation of post-translational modifications detected by mass spectrometry. A second challenge concerned the standardization of acquired native mass spectrometry data, the minimal reporting requirements for these experiments, and the dissemination of these data.
Overall, the 2021 Dagstuhl Seminar on Computational Proteomics was extremely successful as a catalyst for careful yet original thinking about key challenges in the field, and as a means to enable downstream progress by setting important, high impact goals to work on in close collaboration. During this Seminar, new topics for a future Seminar were suggested throughout as well, indicating that this active field will continue to yield novel challenges and opportunities for advanced computational work going forward.
- Sebastian Böcker (Universität Jena, DE) [dblp]
- Viktoria Dorfer (University of Applied Sciences Upper Austria, AT) [dblp]
- Lukas Käll (KTH Royal Institute of Technology - Solna, SE) [dblp]
- Frédérique Lisacek (Swiss Institute of Bioinformatics - Genève, CH) [dblp]
- Lennart Martens (Ghent University, BE) [dblp]
- Magnus Palmblad (Leiden University Medical Center, NL) [dblp]
- Robin Park (Bruker - Rancho Santa Fe, US)
- Daniel Questschlich (University of Oxford, GB)
- Veit Schwämmle (University of Southern Denmark - Odense, DK) [dblp]
- Mathias Wilhelm (TU München - Freising, DE) [dblp]
- Jeffrey Agar (Northeastern University - Boston, US)
- Kiyoko Aoki-Kinoshita (Soka University - Tokyo, JP)
- Marshall Bern (Protein Metrics - Cupertino, US)
- Robert Chalkley (University of California - San Francisco, US)
- Sven Degrove (Ghent University, BE)
- Bernard Delanghe (Thermo Fisher GmbH - Bremen, DE)
- Patrick Emery (Matrix Science Ltd. - London, GB)
- Rebekah Gundry (University of Nebraska - Omaha, US)
- Michael Hoopmann (Institute for Systems Biology - Seattle, US) [dblp]
- Neil Kelleher (Northwestern University - Evanston, US) [dblp]
- Joanna Kirkpatrick (The Francis Crick Institute - London, GB)
- Daniel Kolarich (Griffith University - Southport, AU)
- Rune Linding (HU Berlin, DE) [dblp]
- Nicki Packer (Macquarie University - Sydney, AU)
- Matthew Smith (University of Texas - Austin, US)
- Sabarinath Peruvemba Subramanian (University of Nebraska - Omaha, US)
- Morten Thaysen-Andersen (Macquarie University - Sydney, AU)
- Lilla Turiák (Research Centre for Natural Sciences - Budapest, HU)
- Olga Vitek (Northeastern University - Boston, US) [dblp]
- Christine Vogel (New York University, US) [dblp]
- Dagstuhl Seminar 05471: Computational Proteomics (2005-11-20 - 2005-11-25) (Details)
- Dagstuhl Seminar 08101: Computational Proteomics (2008-03-02 - 2008-03-07) (Details)
- Dagstuhl Seminar 13491: Computational Mass Spectrometry (2013-12-01 - 2013-12-06) (Details)
- Dagstuhl Seminar 15351: Computational Mass Spectrometry (2015-08-23 - 2015-08-28) (Details)
- Dagstuhl Seminar 17421: Computational Proteomics (2017-10-15 - 2017-10-20) (Details)
- Dagstuhl Seminar 19351: Computational Proteomics (2019-08-25 - 2019-08-30) (Details)
- Dagstuhl Seminar 23301: Computational Proteomics (2023-07-23 - 2023-07-28) (Details)
- Machine Learning
- Other Computer Science
- Computational Mass Spectrometry
- Machine Learning