General data
| Credits (SWS) | 12 (10 SWS) |
| Module level | Master |
| Language | German/English |
| Total hours | 360 h |
| Weekly time slot | 2-3 days/week |
| Block part in semester break | to be determined |
Time schedule of the internship
| Feb - Mar 2025 | Kickoff meeting and assignment of projects and teams |
| Apr - Jul 2025 | Division of the project work, interim presentations |
| Jul - Aug 2025 | Blockpart for finalizing the project work, writing the report, and preparing the final presentation |
Requirements and prior knowledge
Bachelor's degree in Bioinformatics. Good programming skills. Interest in data visualization and network medicine. Previous experience in software development is an advantage, but not a must.
Background
Studying the relationship between human health and our microbiota has revealed the importance of the gut microbiome in maintaining human health, kindling interest in exploring the importance of gut microbiome signatures and their impact on our health. For example, microbiota dysbiosis can lead to an overabundance of inflammation-stimulant bacteria like Bacteroides fragilis and Fusobacterium nucleatum (Abed et. al., Bossuet-Greif et al.).
The Collaborative Research Centre (CRC) 1371 aims to determine the functional relevance of gut microbiome signatures and their contribution to human health in a disease-specific manner. With the aim of diverse research groups, the consortium aims to unravel the relevance of changes in the intestinal microbiome and key mechanisms linked to inflammatory bowel disease (IBD) and colorectal cancer.
Motivation
Currently, clinicians lack readily accessible tools to comprehensively analyze and interpret microbiome data, which limits their ability to diagnose dysbiosis and make informed treatment decisions. The ability to discern patients' healthy and unhealthy microbial profiles compared to reference microbial profiles helps clinicians make therapeutic decisions.
The goal of this project is to make the application of microbiome analysis in clinical settings more understandable and practical through several interconnected objectives. As part of the project, you will build up a reference collection of healthy and unhealthy microbial profiles (16S rRNA data). The goal of the project is to develop a clinical report that relates a patient’s microbiome to these reference data to offer insights into potential dysbiosis. You will also work with large language models (LLMs) to extract relevant information from a carefully curated corpus of literature on bacteria-health relationships. To make these tools available to healthcare experts, a user-friendly web interface should be developed that enables clinicians to easily work with the system and its outputs in their practice.
Objectives
- make or prepare reference collections of healthy and unhealthy microbial profiles
- explore or develop scientific methods to compare patient's microbial profiles with a reference collection of profiles (healthy and unhealthy)
- leverage LLMs to explore literature on the relationship of bacteria and health
- corpus preparation
- setting up a Retrieval-Augmented Generation (RAG) LLM process to retrieve targeted information from the corpus
- develop a web interface for clinicians to interact with the application
Tasks
- Familiarisation with health-related microbiome studies and the available dataset.
- Integrating comparison data (healthy cohort).
- Implementing algorithms for analyzing and visualizing microbial profiles.
- Validating and testing the application using provided datasets.
- Designing the architecture and interface of the clinical application (e.g., web app or desktop software).
- Documenting and presenting the results.