Current Open Positions

Internship

1. Comparing Molecular Representations for Predicting Biological Responses

Project description: Modern drug discovery increasingly depends on machine learning models. These models can convert the structure of a molecule into a useful numerical representation for biological predictions. In this project, the student will explore and compare several common chemical representations, including classical fingerprints, SMILES-based language models, graph neural networks (GNNs), and modern pretrained foundation models. The goal is to investigate how these different representations capture chemical information and how well they apply to biological tasks like predicting transcriptional responses, drug effects, or other cell based readouts. The rotation offers a broad introduction to molecular machine learning by mixing intuitive concepts with practical computational experimentation. Throughout the rotation, the student will test various representations on real biological datasets, such as LINCS L1000 perturbation profiles. They will assess their performance using standardized metrics and fair validation methods. The student will gain insight into how chemical structure relates to cellular traits and which types of models work best for new or unseen chemical structures. By the end of the project, the student will have hands-on experience designing controlled machine learning experiments, analyzing high dimensional data, and understanding the strengths and weaknesses of current chemical representations.

Methods/ Techniques: The student will learn data preparation and curation involve handling SMILES strings, molecular graphs, fingerprints, and biological readouts. Classical cheminformatics tools include RDKit for ECFP fingerprints and descriptor calculation. In deep learning part, we will try to train and evaluate models like MLPs, SMILES Transformers, and GNNs. Using pretrained models involves loading and fine-tuning chemical foundation models, such as Chemeleon (different models will be tested based on modular structure of project and time limitations!). Moderate knowledge of Python is required. Basic knowledge of machine learning would be beneficial.

2. Exploratory Analysis of L1000 Gene Expression Signatures for Perturbation Mapping

Project Description: The L1000 dataset from the LINCS program includes gene-expression signatures from human cell lines exposed to thousands of chemical and genetic changes. Researchers widely use this data to study mechanisms of action, drug similarities, and cellular responses. This project focuses on traditional computational analysis of the L1000 dataset. The goal is to explore the structure and variability of the dataset, identify technical and biological variations, and prepare high-quality processed representations for further perturbation modeling, such as clustering, mechanism of action (MoA) prediction, or machine-learning-based embeddings. The student will conduct quality control, normalization, and feature analysis of landmark and inferred genes. They will also examine batch effects, dose-response patterns, and cell-type specificity. This project will offer crucial biological and computational insights that form the groundwork for any future deep-learning project.

Methods/ Techniques: The student will learn L1000 data handling, including metadata, batch structure, replicates, and landmarks. Moreover, quality control of gene expression data, normalization techniques, dimensionality reduction methods such as PCA and UMAP are included. MoA enrichment analysis will be conducted. In addition, student will learn Connectivity Map concepts, including similarity scoring and cosine similarity. Basic knowledge of Python is required for the main part of project.

Contact: 

PhD Student: Fazel Amirvahedi Bonab: fazel.amirvahedi(at)bioinf.med.uni-goettingen.de 

Prof. Dr. Michael Altenbuchinger: michael.altenbuchinger@bioinf.med.uni-goettingen.de

AI and Multi Omics Data: We are looking for an intern, student research assistant, or master’s thesis candidate

We are looking for an intern, student research assistant, or master’s thesis candidate (specific details will depend on the successful applicant) from Georg-August-Universität Göttingen <https://www.linkedin.com/company/-university-of-goettingen/> . The project focuses on an ethical question approached using computer science techniques—most likely involving graph algorithms.

Your profile:

- Strong interest in addressing ethical questions

- Excellent analytical thinking skills

- Solid knowledge of (graph) algorithms and data structures

- Solid programming skills

The ideal candidate will have a computer science background and a passion for contributing to the greater good of society!

For clarification: unlike most projects in our group, this one does not involve biological, medical, pharmaceutical, or life science questions.

How to apply:

Please send your application to me and Daniela Großmann. Your application should include:

- 2–3 sentences explaining why you are a good fit for this position

- Your CV

Submitting a transcript of records is highly encouraged.

More information about the position will be shared upon request!

Thesis offers

In our department five working groups with different focuses are working in the field of knowledge management, data analysis
generegulation, image processing and medical data science.  
We offer places both for students of biological or medical degree programmes, as well as from computer science and chemistry.

In which areas you can successfully work, depends on motivation, previous knowledge and the planned duration of your work.
 

PhD and Dr.rer. nat.
We are currently not actively recruting, but initiative applications from highly motivated students or postdocs are welcome. Please send your application including a letter of motivation, CV, and certificates to Ms. Daniela Großmann.

Medical doctoral thesis
Depending on you background and interest we always try to find a topic within the framework of our research and projects for a medical doctoral thesis. If you are intersted in performing your thesis in our department, please contact our project manager Ms. Daniela Großmann. 
 

Student assistant, practical course, Bachelor and Master theses 

Depending on your field of study, please read our prerequisites/recommendations on our teaching page in advance. Within the scope of our projects, topics for final thesis (Bachelor, Master, PhD) in the Department of Medical Bioinformatics can always occur or if you as student are interested to work in the bioinformatic field (student assistant) please contact Daniela Großmann with some information about your background and motivation.

Also check out the topics and offers of our subgroups under Research and under Projects.