Joint curATime workshop:
T-Rex based high-dimensional Variable Selection with False Discovery Rate Control.
2024/07/23 by fs
As part of the curATime project, supported by the Cluster4Future initiative of the Federal Ministry of Education and Research, a workshop of the cluster project entitled “T-Rex based high-dimensional Variable Selection with False Discovery Rate Control” took place. It was organized by the research groups for Robust Data Science and Signal Processing at TU Darmstadt in collaboration with the Center for Thrombosis and Hemostasis (CTH) of the University Medical Center Mainz at the University Medical Center Mainz.
The workshop started with an in-depth theoretical lecture on high-dimensional learning and variable selection using T-Rex algorithms. This was followed by a lecture on medical data structures, which focused in particular on the data types relevant to the CTH, such as heart rate variability (HRV), lipids, proteins, RNA, methylation and antibodies. The participants discussed their correlation structures and developed concepts for adapting the T-Rex framework to the specific requirements of medical data.
A practical demo presentation on the application of the T-Rex software provided participants with valuable insights, concrete examples and a practical tutorial for self-application. The workshop was enriched by lively technical discussions and excellent catering in a relaxed atmosphere. All participants were enthusiastic and the researchers are looking forward to using the methods in their curATime projects and verifying their research results with the innovations of the T-Rex framework.
The T-Rex software package is already proving very popular and had already been downloaded over 11,000 times from the CRAN open source directory by July 2024.
