Research
Projects move from exploring to validating to published.
01
01
Using genetic and clinical data to describe biologically meaningful groups of acute leukemia, with an emphasis on acute myeloid leukemia and large collaborative cohorts.
02
Data-driven study of graft-versus-host disease, infection, immune reconstitution, and related complications after hematopoietic cell transplantation.
03
Learning from data already collected in hematology care, including laboratory results, pathology reports, and transfusion records, without new specialized assays.
04
Reproducible methods and public software for computational hematology, including the aGVHD grading calculator at gvhd.online.
05
Published tools for decisions in hematology care, including data-driven grading of acute graft-versus-host disease.
06
Making cancer data comparable across hospitals and consortia, including HARMONY Alliance cohorts and IKIM work on structured representations of pathology and clinical text.
02
Public tools and code. The aGVHD calculator belongs with the 2023 Nature Communications grading paper.
Data-driven aGVHD grading · gvhd.online · All public software
03
Programs with a public footprint in papers, abstracts or tools.
Validating
Within the HARMONY Alliance, the group is developing and internationally validating an unsupervised genomic classification of AML. This work has been presented in the EHA 2025 plenary session and as an ASH 2025 oral abstract. It is not yet represented here as a peer-reviewed article.
Cancer genomics and leukemia classification · HARMONY Alliance
Published
The group led a multi-center effort to rebuild acute GVHD grading from organ-stage data. The 2023 Nature Communications paper is accompanied by a public calculator (gvhd.online) and source code.
Transplant outcomes and complications · Multi-center German transplant centers as described in the paper
Recently published
A 2026 Nature Communications paper reports international testing and refinement of algorithms that predict acute leukemia subtypes from routine laboratory data. Related commentary appears in Lancet Digital Health (2024).
AI and real-world clinical data · International hospital network described in the paper
Published
A series of papers from the group examines CMV kinetics, immune reconstitution and relapse risk after HCT, including work led by Saskia Leserer.
Transplant outcomes and complications · Clinical HCT cohorts as described in the papers
Active
This program brings together HARMONY Alliance work on harmonized hematologic malignancy data and IKIM work on turning real-world pathology reports and clinical text into structured representations — including automated tumor ICD coding and clinical embeddings for retrieval-augmented generation.
Cancer data interoperability and harmonization · HARMONY Alliance
Active
The group writes and speaks about how AI can support transplant care and complication management, including a 2025 review with lab co-authors and a 2026 review on big data and AI in transfusion medicine.
AI and real-world clinical data
Active
The lab site lists the PathRoClus Consortium, supported by EHA. Public materials describe a federated platform for hematological malignancies.
Cancer genomics and leukemia classification · PATHroclus consortium
Active
Supported by the José Carreras Leukemia Foundation (DJCLS), this program uses hierarchical Dirichlet mixture models and other data-driven methods to investigate measurable residual disease in AML. It is a collaboration with HARMONY / IBSAL in Salamanca and FIDIS in Santiago de Compostela.
Cancer genomics and leukemia classification · HARMONY Alliance
Active
Building on earlier work on ATLG dosing after unrelated HCT, the group has presented further pharmacokinetics modeling toward individualized dosing. Public description here is limited to conference-abstract level; no unpublished results are given.
Transplant outcomes and complications
04
Published
XplOit was a BMBF i:DSem consortium (2016–2021) coordinated by Fraunhofer IBMT. It built a platform for integrating clinical data and developing prediction models after allogeneic stem cell transplantation. University Hospital Essen contributed clinical expertise and data. This research group did not lead the consortium. A 2022 paper with Amin Turki as last author reports machine-learning prediction of mortality and CMV reactivation after transplantation.
AI and real-world clinical data · XplOit consortium