27 August 2024 in Nature Machine Intelligence
Cellular phenotypic heterogeneity is an important hallmark of many biological processes and understanding its origins remains a substantial challenge. This heterogeneity often reflects variations in the chromatin structure, influenced by factors such as viral infections and cancer, which dramatically reshape the cellular landscape. To address the challenge of identifying distinct cell states, we developed artificial intelligence of the nucleus (AINU), a deep learning method that can identify specific nuclear signatures at the nanoscale resolution. AINU can distinguish different cell states based on the spatial arrangement of core histone H3, RNA polymerase II or DNA from super-resolution microscopy images. With only a small number of images as the training data, AINU correctly identifies human somatic cells, human-induced pluripotent stem cells, very early stage infected cells transduced with DNA herpes simplex virus type 1 and even cancer cells after appropriate retraining. Finally, using AI interpretability methods, we find that the RNA polymerase II localizations in the nucleoli aid in distinguishing human-induced pluripotent stem cells from their somatic cells. Overall, AINU coupled with super-resolution microscopy of nuclear structures provides a robust tool for the precise detection of cellular heterogeneity, with considerable potential for advancing diagnostics and therapies in regenerative medicine, virology and cancer biology.
29 August 2025 in Cellular and Molecular Life Sciences
The 3D organization of the genome plays a critical role in regulating gene expression, maintaining cellular identity, and mediating responses to environmental cues. Advances in super-resolution microscopy and genomic technologies have enabled unprecedented insights into chromatin architecture at nanoscale resolution. However, the complexity and volume of data generated by these techniques necessitate innovative computational strategies for effective analysis and interpretation. In this review, we explore the transformative role of deep learning in the analysis of 3D genome organization, highlighting how deep learning models are being leveraged to enhance image reconstruction, segmentation, and dynamic tracking in chromatin research. We provide an overview of deep learning-enhanced methodologies that significantly improve spatial and temporal resolution of images, with a special focus on single-molecule localization microscopy. Furthermore, we discuss deep learning’s contribution to segmentation accuracy, and its application in single-particle tracking for dissecting chromatin dynamics at the single-cell level. These advances are complemented by frameworks that enable multimodal integration and interpretability, pushing the boundaries of chromatin biology into clinical diagnostics and personalized medicine. Finally, we discuss emerging clinical applications where deep learning models, based on chromatin imaging, aid in disease stratification, drug response prediction, and early cancer detection. We also address the challenges of data sparsity, model interpretability and propose future directions to decode genome function with higher precision and impact.