Ibarra-Arellano Miguel A, Caprio Lindsay A, Hada Aroj, Stotzem Niklas, Cai Luke L, Shah Shivem B, Walsh Zachary H, Melms Johannes C, Wünneman Florian, Bestak Kresimir, Mansaray Ibrahim, Izar Benjamin, Schapiro Denis
Institute for Computational Biomedicine, Heidelberg University, Faculty of Medicine, Heidelberg University Hospital, Heidelberg, Germany.
Department of Medicine, Division of Hematology/Oncology, and Herbert Irving Comprehensive Cancer Center, Columbia University Irving Medical Center, Columbia University Vagelos College of Physician and Surgeons, New York, NY, USA.
Commun Biol. 2025 Mar 4;8(1):361. doi: 10.1038/s42003-025-07796-4.
Chromosomal instability (CIN) is a hallmark of cancer that drives metastasis, immune evasion and treatment resistance. CIN may result from chromosome mis-segregation errors and excessive chromatin is frequently packaged in micronuclei (MN), which can be enumerated to quantify CIN. The assessment of CIN remains a predominantly manual and time-consuming task. Here, we present micronuclAI, a pipeline for automated and reliable quantification of MN of varying size and morphology in cells stained only for DNA. micronuclAI can achieve close to human-level performance on various human and murine cancer cell line datasets. The pipeline achieved a Pearson's correlation of 0.9278 on images obtained at 10X magnification. We tested the approach in otherwise isogenic cell lines in which we genetically dialed up or down CIN rates, and on several publicly available image datasets where we achieved a Pearson's correlation of 0.9620. Given the increasing interest in developing therapies for CIN-driven cancers, this method provides an important, scalable, and rapid approach to quantifying CIN on images that are routinely obtained for research purposes. We release a GUI-implementation for easy access and utilization of the pipeline.
染色体不稳定（CIN）是癌症的一个标志，它会驱动转移、免疫逃逸和治疗抗性。CIN可能源于染色体错分离错误，并且过量的染色质经常被包装在微核（MN）中，微核数量可用于量化CIN。CIN的评估仍然主要是一项人工且耗时的任务。在此，我们展示了micronuclAI，这是一种用于在仅对DNA染色的细胞中自动且可靠地量化不同大小和形态微核的流程。micronuclAI在各种人类和小鼠癌细胞系数据集上能够达到接近人类水平的性能。该流程在10倍放大倍数下获得的图像上实现了0.9278的皮尔逊相关性。我们在基因上调或下调CIN率的同基因细胞系中测试了该方法，并在几个公开可用的图像数据集上进行了测试，在这些数据集上我们实现了0.9620的皮尔逊相关性。鉴于开发针对CIN驱动癌症疗法的兴趣日益增加，该方法为在常规用于研究目的的图像上量化CIN提供了一种重要、可扩展且快速的方法。我们发布了一个图形用户界面（GUI）实现，以便于访问和使用该流程。
