Refine
H-BRS Bibliography
- yes (5)
Departments, institutes and facilities
Document Type
- Article (3)
- Part of a Book (1)
- Conference Object (1)
Language
- English (5)
Keywords
- CDH1 (1)
- E-cadherin (1)
- Familial glioma (1)
- Oligodendroglioma (1)
- Whole-genome sequencing (1)
- β-catenin (1)
The genetic basis of brain tumor development is poorly understood. Here, leukocyte DNA of 21 patients from 15 families with ≥ 2 glioma cases each was analyzed by whole-genome or targeted sequencing. As a result, we identified two families with rare germline variants, p.(A592T) or p.(A817V), in the E-cadherin gene CDH1 that co-segregate with the tumor phenotype, consisting primarily of oligodendrogliomas, WHO grade II/III, IDH-mutant, 1p/19q-codeleted (ODs). Rare CDH1 variants, previously shown to predispose to gastric and breast cancer, were significantly overrepresented in these glioma families (13.3%) versus controls (1.7%). In 68 individuals from 28 gastric cancer families with pathogenic CDH1 germline variants, brain tumors, including a pituitary adenoma, were observed in three cases (4.4%), a significantly higher prevalence than in the general population (0.2%). Furthermore, rare CDH1 variants were identified in tumor DNA of 6/99 (6%) ODs. CDH1 expression was detected in undifferentiated and differentiating oligodendroglial cells isolated from rat brain. Functional studies using CRISPR/Cas9-mediated knock-in or stably transfected cell models demonstrated that the identified CDH1 germline variants affect cell membrane expression, cell migration and aggregation. E-cadherin ectodomain containing variant p.(A592T) had an increased intramolecular flexibility in a molecular dynamics simulation model. E-cadherin harboring intracellular variant p.(A817V) showed reduced β-catenin binding resulting in increased cytosolic and nuclear β-catenin levels reverted by treatment with the MAPK interacting serine/threonine kinase 1 inhibitor CGP 57380. Our data provide evidence for a role of deactivating CDH1 variants in the risk and tumorigenesis of neuroepithelial and epithelial brain tumors, particularly ODs, possibly via WNT/β-catenin signaling.
Development and Validation of a Rapid and Reliable Method for TPMT Genotyping using real-time PCR
(2012)
Deployment of modern data-driven machine learning methods, most often realized by deep neural networks (DNNs), in safety-critical applications such as health care, industrial plant control, or autonomous driving is highly challenging due to numerous model-inherent shortcomings. These shortcomings are diverse and range from a lack of generalization over insufficient interpretability and implausible predictions to directed attacks by means of malicious inputs. Cyber-physical systems employing DNNs are therefore likely to suffer from so-called safety concerns, properties that preclude their deployment as no argument or experimental setup can help to assess the remaining risk. In recent years, an abundance of state-of-the-art techniques aiming to address these safety concerns has emerged. This chapter provides a structured and broad overview of them. We first identify categories of insufficiencies to then describe research activities aiming at their detection, quantification, or mitigation. Our work addresses machine learning experts and safety engineers alike: The former ones might profit from the broad range of machine learning topics covered and discussions on limitations of recent methods. The latter ones might gain insights into the specifics of modern machine learning methods. We hope that this contribution fuels discussions on desiderata for machine learning systems and strategies on how to help to advance existing approaches accordingly.