A novel artificial intelligence-driven methodology facilitates the reconstruction of spatial data concerning the original positioning of immune cells within an organ, even after their extraction and individual examination. Researchers affiliated with the University Hospital Bonn (UKB) and the University of Bonn have achieved this by leveraging the transcriptome, which encompasses all messenger RNA transcripts generated by genes within a cell at a specific juncture. This groundbreaking work, detailing the new MERLIN algorithm, has been disseminated in the esteemed journal Advanced Science.
Understanding the transformation of immune cells and their role in disease pathogenesis within organs presents a significant challenge. The advent of single-cell RNA sequencing technology has profoundly advanced immunological investigations, enabling the identification of active genes within discrete immune cells. However, as Professor Christian Kurts, Director of the Institute for Molecular Medicine and Experimental Immunology at the UKB and a participant in the ImmunoSensation3 Cluster of Excellence and Research Area “Life & Health,” explains, “The isolation of cells inherently leads to the forfeiture of their original organ location. In intricately organized organs like the kidney or brain, this locational context is paramount for deciphering states of health and illness.”
MERLIN Unlocks the Informational Reservoir of Immune Cells
Our findings indicate that macrophages retain a molecular imprint of their indigenous microenvironment. Post-isolation, their gene expression profiles continue to reflect the specific region of the kidney or brain from which they were derived. MERLIN effectively restores access to this crucial information.”
Junping Yin, lead author of the research
MERLIN emerged from a convergence of expertise in immunology, nephrology, and bioinformatics. This sophisticated algorithm employs machine learning techniques to discern distinct patterns in gene activity, which are demonstrably influenced by localized tissue conditions such as hypoxia or variations in salt concentration.
“From a bioinformatics standpoint, ensuring MERLIN’s training on multiple independent data repositories was a critical design element,” states Jian Li, the senior author and a bioinformatician. “This approach enables the system to assimilate authentic biological signals, thereby allowing its application to entirely novel or previously analyzed datasets.”
The research team successfully demonstrated MERLIN’s efficacy not only in murine models but also in accurately predicting the spatial origin of macrophages—a class of substantial, specialized leukocytes—within human renal tissue samples. Furthermore, the methodology was adeptly adapted to the neural context, facilitating the successful reconstruction of the positions of microglia, the resident immune cells of the brain.
MERLIN Illuminates Kidney Disease Pathologies
The application of MERLIN to the study of renal ailments holds particular promise. Through the analysis of pre-existing datasets pertaining to inflammation, sepsis, post-transplant ischemia-reperfusion injury, and diabetic nephropathy, MERLIN corroborated established disease mechanisms while simultaneously unveiling novel insights into region-specific immunological responses and the efficacy of therapeutic interventions. Professor Christian Kurts, a senior author, underscores the significance of this development, stating, “This represents a substantial leap forward for nephrology. We are now able to discern that the nature of immune responses and the impact of pharmaceutical agents are profoundly contingent upon the specific anatomical locality within the kidney, a phenomenon readily observed in clinical practice.”
This investigation was undertaken at the UKB, supported by the ImmunoSensation3 Cluster of Excellence and the TRA “Life & Health” initiative at the University of Bonn, both of which champion interdisciplinary research focused on the immune system. The study also underscores the robust international and national collaborations forged with researchers in Wuhan (China), the University Medical Center Hamburg-Eppendorf, and LMU Munich.
“MERLIN inaugurates a new paradigm in single-cell research,” concludes Junping Yin. “It empowers us to revisit existing datasets and attain a far more granular comprehension of disease etiologies.”
Yin, J., et al. (2026) Predicting Macrophage Spatial Localization from Single-Cell Transcriptomes to Uncover Disease Mechanisms. Advanced Science. DOI: 10.1002/advs.202410924. https://advanced.onlinelibrary.wiley.com/doi/full/10.1002/advs.202410924
