Postdoc (m/w/d)

Heidelberg University Hospital; Heidelberg; On-site; EUR 55,000 - 75,000; Full time

Job summary

Heidelberg University Hospital invites applications for a Postdoctoral Researcher in Translational AI in Neuropathology (m/f/d). The role centers on developing foundation and multimodal models that integrate histomorphology, DNA methylation, sequencing, spatial and clinical data to address unmet clinical needs such as risk stratification and therapy response.

You will work closely with neuropathologists and an interdisciplinary team, contribute to prospective validation, and build reproducible

Responsibilities

• Develop and adapt foundation and multimodal models integrating histomorphology, DNA methylation, sequencing, spatial and clinical data.
• Establish reproducible training and inference pipelines and collaborate with neuropathologists.

Job description

The University Hospital Heidelberg is one of the major healthcare centers in Germany. Our objective is the development of innovative diagnostics and therapies as well as their quick implementation for the patient. With about 14,000 employees in more than 50 specialized clinical departments with almost 2,500 beds, about 80,000 patients in part-time and full-time inpatient treatment as well as 1,000,000 patients in ambulant treatment are medicated each year.

Postdoctoral Researcher - Translational AI in Neuropathology (m/f/d)

The Department of Neuropathology is one of the world's leading reference centres for CNS tumour diagnostics. We co-developed the DNA methylation-based classifier for CNS tumours and continue to expand it (Capper et al., 2018; Sill et al., 2025), and we combine fully digitised histopathology, genome-wide methylation profiling, panel and long-read sequencing and spatial multi-omics with deeply annotated outcome data across tens of thousands of cases (Ritter et al., Nat Cancer 2026; Maas et al., Nat Genetics 2026). Our recent work includes Hetairos, a histology-based AI model that predicts CNS tumour methylation subtypes directly from H&E slides (Jin et al., Nat Cancer 2026), and we are the coordinating site of the EUcanAI (European Unified Cancer diagnostics and treatment AI agent) consortium, an EIC Pathfinder Challenge project on generative AI agents for cancer diagnosis and treatment.

We are now expanding our activities in translational artificial intelligence and are looking for an ambitious postdoctoral researcher who wants to grow into the leadership of an independent research group. You will develop foundation and multimodal models that integrate histomorphology, DNA methylation, sequencing, spatial and clinical data, to address unmet clinical needs (risk stratification, therapy response, identification and characterization of rare entities, AI-supported process optimization) and take models from retrospective proof of concept through prospective validation towards certified diagnostic use.

As a key member of our interdisciplinary team of physicians, biologists and computer scientists, you will work directly alongside the neuropathologists who sign out the diagnoses and with the data your models learn from. We are a key part of the Heidelberg (neuro)-oncology ecosystem and enjoy close collaboration with diverse clinical and technical departments on campus as well as active partnerships with renowned international institutions.

We will actively support you in building your own scientific profile, in acquiring independent third-party funding and in your habilitation - with the clear perspective of establishing your own research group within the department.

Selected recent work from the department

• Jin D, et al. Hetairos is a histology-based artificial intelligence model for predicting central nervous system tumor methylation subtypes. Nat Cancer. 2026;7:884-898. doi:10.1038/s43018-026-01186-3
• Sill M, et al. Advancing CNS tumor diagnostics with expanded DNA methylation-based classification. Cancer Cell. 2025;44:340-354. doi:10.1016/j.ccell.2025.11.002
• Maas SLN, et al. A microenvironment-determined risk continuum refines subtyping in meningioma and reveals determinants of machine learning-based tumor classification. Nat Genet. 2026;58:341-354. doi:10.1038/s41588-025-02475-w
• Capper D, et al. DNA methylation-based classification of central nervous system tumours. Nature. 2018;555:469-474. doi:10.1038/nature26000

• Job-ID: V000016389
• Field of application: Neuropathologie
• Location: Heidelberg
• Start date: 01.12.2026
• Job Category: Science and teaching
• Working hours: full-time (100%)
• Limitation: Temporary (3 Jahre)
• Contract: TV-L

As a key member of our interdisciplinary team, you will:

• Develop and adapt foundation and multimodal models integrating histomorphology, -DNA methylation, sequencing, spatial and clinical data Establish reproducible training and inference pipelines o...