Group members
Titel | Name | Position/Tätigkeiten | Kontakt |
---|---|---|---|
Elke Schubert | Office Assistant | elke.schubert(at)dzne.de | |
Dr. | Marie Oestreich | Postdoc | marie.oestreich(at)dzne.de |
Shubhi Ambast | PhD Student | shubhi.ambast(at)dzne.de | |
Charles Mwangi Kaumbutha | PhD Student | charlesmwangi.kaumbutha(at)dzne.de | |
Karola Mai | PhD Student | karola.mai(at)dzne.de | |
Dayoung Lee | Master Student | dayoung.lee(at)dzne.de | |
Philip Horvat | Labrotation Student | philiplucien.horvat(at)dzne.de |
Shubhi Ambast
![](/fileadmin/_processed_/a/a/csm_shubhi_c5a32ed447.jpg)
I am interested in applying deep learning methods, particularly graph neural networks, to unravel complex biological insights from datasets such as those generated by single-cell transcriptomics. I am also intrigued by the use of explainable AI to interpret and validate my research, providing a deeper understanding of the underlying biological mechanisms in healthcare research.
Karola Mai
![](/fileadmin/_processed_/0/e/csm_karola_16ecd14736.jpg)
My research focuses on the PriSyn project, where I evaluate the biological plausibility of generated synthetic scRNAseq data and work on integrating scRNAseq data with genotype information. Additionally, leveraging my hands-on experience in flow cytometry, I now use machine learning methods to enhance the standardisation and scalability of flow cytometry data analysis.
Charles Mwangi Kaumbutha
![](/fileadmin/_processed_/6/8/csm_charles_129828a078.jpg)
My research interest is in Meta-Data Analysis particularly addressing missing data. Currently, my primary focus is examining evaluation metrics for imputation methods, to ensure quality downstream machine learning tasks such as classification.
Marie Oestreich
![](/fileadmin/_processed_/a/6/csm_marie_dd074c40b4.jpg)
My research focuses on chemoinformatics, specifically on developing deep generative models that propose novel small molecules to speed up the drug development process. In this context, I further investigate chemically-informed embedding strategies for molecules to maximise chemical information content during training of downstream deep learning models.