A new AI-assisted pipeline called MAGIC enables researchers to analyze nearly 100,000 cells in under a day, letting them study how often chromosomal abnormalities arise during cell division and how factors like p53 mutations influence these rates.

MAGIC automates detection of micronuclei—DNA-containing structures linked to chromosomal abnormalities—using an autonomous imaging system and a photoconvertible dye to permanently tag affected cells for deeper analysis.

Developed at EMBL Heidelberg, MAGIC combines automated microscopy, single-cell sequencing, and machine learning to study the origins and rates of chromosomal abnormalities in normal cells during mitosis.

The work emerges from an interdisciplinary collaboration among computer vision, robotics, genomics, and microscopy teams at EMBL Heidelberg and partners, illustrating how AI can accelerate single-cell analysis and open new diagnostic and therapeutic avenues.

The project reflects a multi-institutional effort involving EMBL groups (ALMF and Pepperkok Team), EMBL-EBI, and DKFZ, underscoring a broad collaboration to advance AI-assisted genomics and imaging.

Beyond cancer, the platform is adaptable to other fields, with algorithms that can be tuned to various cellular features, potentially benefiting neuroscience, immunology, and other areas needing visual phenotyping linked to molecular data.

Lead researchers, including EMBL senior scientist Jan Korbel, frame the technology as cutting-edge AI-driven biology with the potential to reveal fundamental cellular traits and origins of disease.

MAGIC demonstrated versatility by showing it can be trained to detect different visually discriminable cellular features beyond micronuclei, speeding discoveries across biology.

Collaborations across EMBL Heidelberg, EMBL-EBI, and DKFZ highlight MAGIC’s adaptability for detecting a range of visually distinguishable cellular features beyond micronuclei.

The approach is adaptable to identify other visually discriminable cellular features, broadening potential applications in biology.

The study also investigates contributors to chromosomal abnormalities, including the presence and location of double-stranded DNA breaks within chromosomes.

The research, published in Nature as Origins of chromosome instability unveiled by coupled imaging and genomics, represents collaboration among EMBL Heidelberg groups (Korbel, ALMF, Pepperkock Team), EMBL-EBI, and DKFZ.