EkoEstMed

Developments in clinical-oriented decision support for high-throughput data in personalized medicine

About the project

Nowadays, precision medicine is becoming more and more of a reality and is moving further and further into clinical practice. Unique characteristics of The goal is to identify molecular features that may influence treatment recommendations and disease outcome as well as response to treatment. Therefore, we aim to estimate the likelihood of developing a particular type of cancer or the likelihood of recurrence and ultimately life expectancy.patients are identified to improve the efficacy and accuracy of their treatments. Ambitious initiatives such as the medical informatics initiative (MII) actively support biomedical research in Germany to improve individualized patient care. It relies heavily on the ever-growing amount of health-related data and constantly evolving technologies. However, due to a lack of analytical methods and standardization and harmonization processes, only part of the existing data is used for clinical recommendations. Thus, the goal is to prepare and present the complex data in an understandable form through the development of innovative methods, such as machine learning (artificial intelligence) algorithms, and analysis processes, while making them available for exchange in an efficient and usable manner. In addition, efficient IT infrastructures as well as the development of novel visualization standards will be created to ensure implementation and application in the clinic. This project will not only advance the processes of personalized medicine in the local clinical setting, but also MII and MIRACUM at various sites.

---

Goals

1. Methods for the development, provision and use of omics data

The goal is to develop standardization methods to properly store, annotate, and share the omics/high-throughput data being generated, while also developing visualization methods for preparing and using this high-dimensional data for clinical applications.

2. Integration of multi-omics data

The integration of omics data, which on the one hand originate from different clinical laboratories, and on the other hand can be used from public databases, are usually accompanied by possible batch effects. For this purpose, appropriate methods and approaches of batch effect correction will be developed and applied. At the same time, other co-factors such as gender, age, and environment will be taken into account. In addition, correlation and interaction analyses of the different omics data sets will be applied and further developed, which will finally lead to a score for the assessment of pathogenicity.

3. Development of decision support tools in the clinical context

The goal is to identify molecular features that may influence treatment recommendations and disease outcome as well as response to treatment. Therefore, we aim to estimate the likelihood of developing a particular type of cancer or the likelihood of recurrence and ultimately life expectancy.

---

Lead of the Junior Research Group

---