PhD Researcher / Doctoral Researcher (m/f/d) Bioinformatics & AML

Department of Internal Medicine II, Research Group: Functional Genomics of Acute Leukemia (Heads: Dr. Lorenz Bastian / Prof. Dr. Claudia Baldus)

The position is part of the newly established research project “LUCID-AML – Learning-based Unsupervised Clustering of transcriptomIcs to Detect developmental, functional and prognostic subtypes in AML across age”, funded by the Deutsche José Carreras Leukämie-Stiftung. It is closely linked to our DFG-funded Clinical Research Unit “CATCH-ALL: Towards a Cure for Adults and Children with Acute Lymphoblastic Leukemia”, where clinical and basic researchers work together to define mechanisms of leukemogenesis and treatment resistance across age groups.

Acute myeloid leukemia (AML) affects people of all ages, but its incidence increases steeply in older adults, while survival decreases sharply with age. AML can be classified into transcriptomic groups that reflect cytogenetic lesions, reveal developmental origins and link leukemias to normal hematopoietic states. However, consistent subtype definitions and robustly validated tools for applying them across cohorts are lacking.

The project builds on ALLCatchR and our work on machine-learning frameworks for disease classification and the identification of novel molecular subtypes in acute lymphoblastic leukemia. LUCID-AML will integrate more than 8,000 bulk AML gene-expression profiles from multiple cohorts with single-cell data, molecular genetics, diagnostic classifications, drug-response data and clinical outcomes. The project aims to establish reproducible, gene-expression-based AML subtypes while accounting for age and treatment. Further aims are to define their developmental origins and make the resulting models available as transparent, open-source tools for research and clinical diagnostics. For further reading: Beder et al., HemaSphere, 2023; Bastian et al., Blood, 2024; Beder et al., bioRxiv / HemaSphere, 2026.

Join our team

Join our team as soon as possible. The initial appointment is for three years, with the possibility of extension.

What we offer:

• Remuneration according to the German TV-L E13 pay scale for a 65% position (25 hours per week), subject to the applicable collective agreement requirements; the position offers the opportunity to pursue a PhD
• Access to a large AML transcriptomics resource with comprehensive molecular, functional and clinical annotations across a broad age range
• Close collaboration with computational and clinical scientists, bioinformaticians, experimental hematologists and clinician-scientists, including national and international partners
• Training and research opportunities in large-scale bulk and single-cell transcriptomics, machine learning, R package development and reproducible research, with access to high-performance computing infrastructure
• Additional UKSH employee benefits according to institutional policies: Benefits

Your tasks:

• Harmonize RNA-seq datasets from multiple cohorts and age groups and develop an end-to-end subtype-discovery workflow, including quality control, batch correction, feature selection, manifold learning, subtype definition and stability assessment
• Build calibrated hierarchical machine-learning models that transfer subtype assignments across cohorts and platforms
• Evaluate subtype definitions in relation to molecular genetic findings, WHO/ICC and ELN categories, age, treatment, drug response and clinical outcomes
• Map AML subtypes onto atlases of hematopoietic development, assess their stemness states and validate bulk-derived findings using large-scale single-cell RNA-seq datasets
• Implement and document the best-performing methods in a reproducible R package and contribute to collaborative analyses, manuscripts, presentations and project reporting

Your profile:

• We are looking for an enthusiastic doctoral researcher with a degree (M.Sc. or equivalent) in bioinformatics, computational biology, data science, systems biology, molecular biology or a related quantitative or life-science discipline
• Programming skills in R and/or Python and a solid foundation in statistics or machine learning are required; willingness to work extensively in R is essential
• Experience with bulk or single-cell transcriptomics, cancer genomics, clustering or classification, package development, workflow systems or containers is advantageous but can also be acquired during the project
• We are looking for a focused, collaborative candidate with a strong work ethic, the motivation to develop an in-depth understanding of the project
• The ability to work independently as well as within the group and wider consortium

Please submit your application until 28. October 2026, indicating the reference number 29315.