Refine
Document Type
- Article (4)
- Conference Proceeding (1)
- Doctoral Thesis (1)
- Part of Periodical (1)
Language
- English (7) (remove)
Is part of the Bibliography
- yes (7) (remove)
Keywords
- Clinical predictive modeling (1)
- Cloud Computing (1)
- Forschungsprojekte (1)
- Future SOC Lab (1)
- HiGHmed (1)
- In-Memory Technologie (1)
- Interpretability (1)
- Multicore Architekturen (1)
- Nephrology (1)
- R Shiny (1)
HPI Future SOC Lab
(2015)
Das Future SOC Lab am HPI ist eine Kooperation des Hasso-Plattner-Instituts mit verschiedenen Industriepartnern. Seine Aufgabe ist die Ermöglichung und Förderung des Austausches zwischen Forschungsgemeinschaft und Industrie.
Am Lab wird interessierten Wissenschaftlern eine Infrastruktur von neuester Hard- und Software kostenfrei für Forschungszwecke zur Verfügung gestellt. Dazu zählen teilweise noch nicht am Markt verfügbare Technologien, die im normalen Hochschulbereich in der Regel nicht zu finanzieren wären, bspw. Server mit bis zu 64 Cores und 2 TB Hauptspeicher. Diese Angebote richten sich insbesondere an Wissenschaftler in den Gebieten Informatik und Wirtschaftsinformatik. Einige der Schwerpunkte sind Cloud Computing, Parallelisierung und In-Memory Technologien.
In diesem Technischen Bericht werden die Ergebnisse der Forschungsprojekte des Jahres 2015 vorgestellt. Ausgewählte Projekte stellten ihre Ergebnisse am 15. April 2015 und 4. November 2015 im Rahmen der Future SOC Lab Tag Veranstaltungen vor.
Portal Wissen = Health
(2020)
The Coronavirus pandemic has made it very clear how much health and well-being determine our lives. And that science led the way in this regard could not be ignored. At the University of Potsdam, too, many researchers deal with aspects of health maintenance, whether in nutritional sciences, sports and rehabilitation medicine, biochemistry, or psychology. Their research includes supporting chronically ill children and the professional handling of risks, as you can read in this issue of our magazine.
With the establishment of our seventh faculty, the Faculty of Health Sciences, these and many new medical topics are getting more attention at the University of Potsdam. While in the beginning, the “Brandenburg Health Campus” funded by the federal state of Brandenburg was a virtual network of university and non-university research, it is now getting more points of intersection and, not only since COVID-19, a very practical use and plausible to everyone.
The Faculty of Health Sciences, founded in 2018, is supported by three institutions: the University of Potsdam, the Brandenburg Technical University Cottbus-Senftenberg and the Brandenburg Medical School in Neuruppin. They pursue an interdisciplinary approach that holistically develops teaching, transfers new scientific findings from theory to practice and thus further improves overall medical care in Brandenburg. Their vision of being a central platform of research, teaching, and transfer combines socially relevant issues and existing expertise to align them with the needs of people in Brandenburg and use them to their benefit. This interdisciplinary structure has never been more important to advancing patient-oriented basic research and health care models. An innovative concept that can make Brandenburg a pioneer.
In the meantime, the Faculty of Health Sciences has established 16 new professorships at the supporting universities, which are concerned with medicine and healthy aging, health services research, nursing and rehabilitation sciences, and telemedicine. Cardiology and physiology will play a central role as well. In general, the innovative faculty counts on strong interdisciplinary relationships, for example with nutritional sciences and the digital health department at the Digital Engineering Faculty. The role of digitization and well-prepared data in combating the Coronavirus pandemic can also be read about in this issue.
As usual, the research magazine addresses the full range of research at the university: We introduce historian Dominik Geppert, who deals with the history of unified Germany after 1990 embedded in the tensions created by a context of national unification, European integration, and global networking. In a self-experiment, we explored together with a psycholinguist how to research word-finding disorders. Last but not least, we were able to take part in a trip to Namibia, where ecologists from Potsdam examine wildlife management in the threatened savannah. Let them take you where kudu and springbok live!
Despite advances in machine learning-based clinical prediction models, only few of such models are actually deployed in clinical contexts. Among other reasons, this is due to a lack of validation studies. In this paper, we present and discuss the validation results of a machine learning model for the prediction of acute kidney injury in cardiac surgery patients initially developed on the MIMIC-III dataset when applied to an external cohort of an American research hospital. To help account for the performance differences observed, we utilized interpretability methods based on feature importance, which allowed experts to scrutinize model behavior both at the global and local level, making it possible to gain further insights into why it did not behave as expected on the validation cohort. The knowledge gleaned upon derivation can be potentially useful to assist model update during validation for more generalizable and simpler models. We argue that interpretability methods should be considered by practitioners as a further tool to help explain performance differences and inform model update in validation studies.
Precision oncology is a rapidly evolving interdisciplinary medical specialty. Comprehensive cancer panels are becoming increasingly available at pathology departments worldwide, creating the urgent need for scalable cancer variant annotation and molecularly informed treatment recommendations. A wealth of mainly academia-driven knowledge bases calls for software tools supporting the multi-step diagnostic process. We derive a comprehensive list of knowledge bases relevant for variant interpretation by a review of existing literature followed by a survey among medical experts from university hospitals in Germany. In addition, we review cancer variant interpretation tools, which integrate multiple knowledge bases. We categorize the knowledge bases along the diagnostic process in precision oncology and analyze programmatic access options as well as the integration of knowledge bases into software tools. The most commonly used knowledge bases provide good programmatic access options and have been integrated into a range of software tools. For the wider set of knowledge bases, access options vary across different parts of the diagnostic process. Programmatic access is limited for information regarding clinical classifications of variants and for therapy recommendations. The main issue for databases used for biological classification of pathogenic variants and pathway context information is the lack of standardized interfaces. There is no single cancer variant interpretation tool that integrates all identified knowledge bases. Specialized tools are available and need to be further developed for different steps in the diagnostic process.
Eatomics
(2021)
Quantitative proteomics data are becoming increasingly more available, and as a consequence are being analyzed and interpreted by a larger group of users. However, many of these users have less programming experience. Furthermore, experimental designs and setups are getting more complicated, especially when tissue biopsies are analyzed. Luckily, the proteomics community has already established some best practices on how to conduct quality control, differential abundance analysis and enrichment analysis. However, an easy-to-use application that wraps together all steps for the exploration and flexible analysis of quantitative proteomics data is not yet available. For Eatomics, we utilize the R Shiny framework to implement carefully chosen parts of established analysis workflows to (i) make them accessible in a user-friendly way, (ii) add a multitude of interactive exploration possibilities, and (iii) develop a unique experimental design setup module, which interactively translates a given research hypothesis into a differential abundance and enrichment analysis formula. In this, we aim to fulfill the needs of a growing group of inexperienced quantitative proteomics data analysts. Eatomics may be tested with demo data directly online via https://we.analyzegenomes.com/now/eatomics/or with the user's own data by installation from the Github repository at https://github.com/Millchmaedchen/Eatomics.