Titulo Estágio
Computational biomarker research
Local do Estágio
Roche Diagnostics GmbH, Roche Diagnostics GmbH, Penzberg, Germany
Enquadramento
Early detection of biomarkers is vital for timely diagnosis and treatment in various medical fields, including neurology, oncology, infectious diseases, and cardiology. The use of multi-modal data types, such as e.g. transcriptomics, imaging, and methylation data, offers a promising avenue for biomarker discovery through the application of machine learning methods.
This master thesis primarily focuses on method development for biomarker discovery using machine learning techniques. The objective is to develop and apply innovative algorithms and approaches that can effectively integrate and learn patterns from diverse multi-modal data types. By leveraging machine learning, this research aims to enhance the accuracy, robustness, and efficiency of biomarker identification
Objetivo
The thesis will involve the design and implementation of novel computational methods for data integration and feature selection. These methods will be tailored specifically for analyzing multi-modal data, allowing the extraction of relevant biomarkers that may not be evident from individual data types alone.
The developed methods will be evaluated using benchmark datasets from the public domain. Performance metrics such as accuracy, sensitivity, and specificity will be used to assess the effectiveness of the proposed approaches in identifying biomarker candidates.
While the thesis primarily focuses on method development, the overall goal is to contribute to the field of biomarker discovery by providing novel computational approaches that can advance early diagnosis and improve patient outcomes. By leveraging machine learning techniques and multi-modal data integration, this research endeavors to advance our
understanding of biomarkers and enhance their clinical utility in various medical domains.
Plano de Trabalhos - Semestre 1
Review the state of the art / literature
- Review of relevant public repositories and identification of relevant public domain
studies
- Review the work that has been done within Roche so far
- Familiarisation with the Roche environment
- Familiarisation with relevant data types
- Interviewing of main stakeholders
- Initial workflow design: data pre-processing, data representation, data curation, data visualisation, model validation, definition of test and validation sets
- First prototyping and collection of feedback
- Writing of interim report
Plano de Trabalhos - Semestre 2
- Iterative improvements of the prototype based on stakeholder feedback
- Application of performance metrics on test and validation sets
- Refinement of data processing and fine tuning of models to increase the performance
- Application of the developed framework in a relevant medical field
- Identification of potential biomarker candidates
- Biological contextualising and interpretation of the findings within the field in close cooperation with SMEs
- Writing of the master thesis
Condições
The work will be carried out at Roche Diagnostics in Penzberg, Germany. At Roche the student will have at his disposal all the tools and resources (hardware and software) required.
Observações
About the Roche supervisor:
Title, full name: Manuel Dietrich
E-mail address: manuel.dietrich@roche.com
Phone number: +49 621 759 67356
Academic degree, scientific area, date awarded: Master of Bioinformatics, 2006
Job position: Senior Data Scientist
Orientador
Manuel Dietrich
manuel.dietrich@roche.com 📩