1
|
Ma C, Tan W, He R, Yan B. Pretraining a foundation model for generalizable fluorescence microscopy-based image restoration. Nat Methods 2024; 21:1558-1567. [PMID: 38609490 DOI: 10.1038/s41592-024-02244-3] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [MESH Headings] [Grants] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 07/27/2023] [Accepted: 03/13/2024] [Indexed: 04/14/2024]
Abstract
Fluorescence microscopy-based image restoration has received widespread attention in the life sciences and has led to significant progress, benefiting from deep learning technology. However, most current task-specific methods have limited generalizability to different fluorescence microscopy-based image restoration problems. Here, we seek to improve generalizability and explore the potential of applying a pretrained foundation model to fluorescence microscopy-based image restoration. We provide a universal fluorescence microscopy-based image restoration (UniFMIR) model to address different restoration problems, and show that UniFMIR offers higher image restoration precision, better generalization and increased versatility. Demonstrations on five tasks and 14 datasets covering a wide range of microscopy imaging modalities and biological samples demonstrate that the pretrained UniFMIR can effectively transfer knowledge to a specific situation via fine-tuning, uncover clear nanoscale biomolecular structures and facilitate high-quality imaging. This work has the potential to inspire and trigger new research highlights for fluorescence microscopy-based image restoration.
Collapse
Affiliation(s)
- Chenxi Ma
- School of Computer Science, Shanghai Key Laboratory of Intelligent Information Processing, Fudan University, Shanghai, China
| | - Weimin Tan
- School of Computer Science, Shanghai Key Laboratory of Intelligent Information Processing, Fudan University, Shanghai, China
| | - Ruian He
- School of Computer Science, Shanghai Key Laboratory of Intelligent Information Processing, Fudan University, Shanghai, China
| | - Bo Yan
- School of Computer Science, Shanghai Key Laboratory of Intelligent Information Processing, Fudan University, Shanghai, China.
| |
Collapse
|
2
|
Li X, Hu X, Chen X, Fan J, Zhao Z, Wu J, Wang H, Dai Q. Spatial redundancy transformer for self-supervised fluorescence image denoising. NATURE COMPUTATIONAL SCIENCE 2023; 3:1067-1080. [PMID: 38177722 PMCID: PMC10766531 DOI: 10.1038/s43588-023-00568-2] [Citation(s) in RCA: 1] [Impact Index Per Article: 1.0] [Reference Citation Analysis] [Abstract] [Key Words] [MESH Headings] [Grants] [Track Full Text] [Subscribe] [Scholar Register] [Received: 06/14/2023] [Accepted: 11/07/2023] [Indexed: 01/06/2024]
Abstract
Fluorescence imaging with high signal-to-noise ratios has become the foundation of accurate visualization and analysis of biological phenomena. However, the inevitable noise poses a formidable challenge to imaging sensitivity. Here we provide the spatial redundancy denoising transformer (SRDTrans) to remove noise from fluorescence images in a self-supervised manner. First, a sampling strategy based on spatial redundancy is proposed to extract adjacent orthogonal training pairs, which eliminates the dependence on high imaging speed. Second, we designed a lightweight spatiotemporal transformer architecture to capture long-range dependencies and high-resolution features at low computational cost. SRDTrans can restore high-frequency information without producing oversmoothed structures and distorted fluorescence traces. Finally, we demonstrate the state-of-the-art denoising performance of SRDTrans on single-molecule localization microscopy and two-photon volumetric calcium imaging. SRDTrans does not contain any assumptions about the imaging process and the sample, thus can be easily extended to various imaging modalities and biological applications.
Collapse
Affiliation(s)
- Xinyang Li
- Department of Automation, Tsinghua University, Beijing, China
- Tsinghua Shenzhen International Graduate School, Tsinghua University, Shenzhen, China
- Institute for Brain and Cognitive Sciences, Tsinghua University, Beijing, China
| | - Xiaowan Hu
- Tsinghua Shenzhen International Graduate School, Tsinghua University, Shenzhen, China
| | - Xingye Chen
- Department of Automation, Tsinghua University, Beijing, China
- Institute for Brain and Cognitive Sciences, Tsinghua University, Beijing, China
- Research Institute for Frontier Science, Beihang University, Beijing, China
| | - Jiaqi Fan
- Tsinghua Shenzhen International Graduate School, Tsinghua University, Shenzhen, China
- Department of Electronic Engineering, Tsinghua University, Beijing, China
| | - Zhifeng Zhao
- Department of Automation, Tsinghua University, Beijing, China
- Institute for Brain and Cognitive Sciences, Tsinghua University, Beijing, China
| | - Jiamin Wu
- Department of Automation, Tsinghua University, Beijing, China.
- Institute for Brain and Cognitive Sciences, Tsinghua University, Beijing, China.
- Beijing Key Laboratory of Multi-dimension and Multi-scale Computational Photography (MMCP), Tsinghua University, Beijing, China.
- IDG/McGovern Institute for Brain Research, Tsinghua University, Beijing, China.
| | - Haoqian Wang
- Tsinghua Shenzhen International Graduate School, Tsinghua University, Shenzhen, China.
- The Shenzhen Institute of Future Media Technology, Shenzhen, China.
| | - Qionghai Dai
- Department of Automation, Tsinghua University, Beijing, China.
- Institute for Brain and Cognitive Sciences, Tsinghua University, Beijing, China.
- Beijing Key Laboratory of Multi-dimension and Multi-scale Computational Photography (MMCP), Tsinghua University, Beijing, China.
- IDG/McGovern Institute for Brain Research, Tsinghua University, Beijing, China.
| |
Collapse
|
3
|
Demmerle J, Hao S, Cai D. Transcriptional condensates and phase separation: condensing information across scales and mechanisms. Nucleus 2023; 14:2213551. [PMID: 37218279 PMCID: PMC10208215 DOI: 10.1080/19491034.2023.2213551] [Citation(s) in RCA: 4] [Impact Index Per Article: 4.0] [Reference Citation Analysis] [Abstract] [Key Words] [MESH Headings] [Grants] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 01/10/2023] [Revised: 04/26/2023] [Accepted: 05/10/2023] [Indexed: 05/24/2023] Open
Abstract
Transcription is the fundamental process of gene expression, which in eukaryotes occurs within the complex physicochemical environment of the nucleus. Decades of research have provided extreme detail in the molecular and functional mechanisms of transcription, but the spatial and genomic organization of transcription remains mysterious. Recent discoveries show that transcriptional components can undergo phase separation and create distinct compartments inside the nucleus, providing new models through which to view the transcription process in eukaryotes. In this review, we focus on transcriptional condensates and their phase separation-like behaviors. We suggest differentiation between physical descriptions of phase separation and the complex and dynamic biomolecular assemblies required for productive gene expression, and we discuss how transcriptional condensates are central to organizing the three-dimensional genome across spatial and temporal scales. Finally, we map approaches for therapeutic manipulation of transcriptional condensates and ask what technical advances are needed to understand transcriptional condensates more completely.
Collapse
Affiliation(s)
- Justin Demmerle
- Department of Biochemistry and Molecular Biology, Johns Hopkins Bloomberg School of Public Health, Baltimore, MD, USA
| | - Siyuan Hao
- Department of Biochemistry and Molecular Biology, Johns Hopkins Bloomberg School of Public Health, Baltimore, MD, USA
| | - Danfeng Cai
- Department of Biochemistry and Molecular Biology, Johns Hopkins Bloomberg School of Public Health, Baltimore, MD, USA
- Department of Biophysics and Biophysical Chemistry, Johns Hopkins School of Medicine, Baltimore, MD, USA
- Department of Oncology, Johns Hopkins School of Medicine, Baltimore, MD, USA
| |
Collapse
|
4
|
Levet F. Optimizing Voronoi-based quantifications for reaching interactive analysis of 3D localizations in the million range. FRONTIERS IN BIOINFORMATICS 2023; 3:1249291. [PMID: 37600969 PMCID: PMC10436483 DOI: 10.3389/fbinf.2023.1249291] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Key Words] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 06/28/2023] [Accepted: 07/27/2023] [Indexed: 08/22/2023] Open
Abstract
Over the last decade, single-molecule localization microscopy (SMLM) has revolutionized cell biology, making it possible to monitor molecular organization and dynamics with spatial resolution of a few nanometers. Despite being a relatively recent field, SMLM has witnessed the development of dozens of analysis methods for problems as diverse as segmentation, clustering, tracking or colocalization. Among those, Voronoi-based methods have achieved a prominent position for 2D analysis as robust and efficient implementations were available for generating 2D Voronoi diagrams. Unfortunately, this was not the case for 3D Voronoi diagrams, and existing methods were therefore extremely time-consuming. In this work, we present a new hybrid CPU-GPU algorithm for the rapid generation of 3D Voronoi diagrams. Voro3D allows creating Voronoi diagrams of datasets composed of millions of localizations in minutes, making any Voronoi-based analysis method such as SR-Tesseler accessible to life scientists wanting to quantify 3D datasets. In addition, we also improve ClusterVisu, a Voronoi-based clustering method using Monte-Carlo simulations, by demonstrating that those costly simulations can be correctly approximated by a customized gamma probability distribution function.
Collapse
Affiliation(s)
- Florian Levet
- CNRS, Interdisciplinary Institute for Neuroscience, IINS, UMR 5297, University of Bordeaux, Bordeaux, France
- CNRS, INSERM, Bordeaux Imaging Center, BIC, UAR3420, US 4, University of Bordeaux, Bordeaux, France
| |
Collapse
|
5
|
Cosma MP, Neguembor MV. The magic of unraveling genome architecture and function. Cell Rep 2023; 42:112361. [PMID: 37059093 DOI: 10.1016/j.celrep.2023.112361] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Key Words] [MESH Headings] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 12/06/2022] [Revised: 02/20/2023] [Accepted: 03/22/2023] [Indexed: 04/16/2023] Open
Abstract
Over the last decades, technological breakthroughs in super-resolution microscopy have allowed us to reach molecular resolution and design experiments of unprecedented complexity. Investigating how chromatin is folded in 3D, from the nucleosome level up to the entire genome, is becoming possible by "magic" (imaging genomic), i.e., the combination of imaging and genomic approaches. This offers endless opportunities to delve into the relationship between genome structure and function. Here, we review recently achieved objectives and the conceptual and technical challenges the field of genome architecture is currently undertaking. We discuss what we have learned so far and where we are heading. We elucidate how the different super-resolution microscopy approaches and, more specifically, live-cell imaging have contributed to the understanding of genome folding. Moreover, we discuss how future technical developments could address remaining open questions.
Collapse
Affiliation(s)
- Maria Pia Cosma
- Medical Research Institute, Guangdong Provincial People's Hospital (Guangdong Academy of Medical Sciences), Southern Medical University, 106 Zhongshan Er Road, Yuexiu District, 510080 Guangzhou, China; Centre for Genomic Regulation (CRG), The Barcelona Institute of Science and Technology, 08003 Barcelona, Spain; Universitat Pompeu Fabra (UPF), Dr Aiguader 88, 08003 Barcelona, Spain; ICREA, Pg. Lluís Companys 23, 08010 Barcelona, Spain.
| | - Maria Victoria Neguembor
- Centre for Genomic Regulation (CRG), The Barcelona Institute of Science and Technology, 08003 Barcelona, Spain.
| |
Collapse
|
6
|
Valenta H, Quiblier N, Laghi V, Cabriel C, Riti J. Latest trends in bioimaging and building a proactive network of early-career young scientists around bioimaging in Europe. Biol Open 2022; 11:286144. [PMID: 36541651 PMCID: PMC9793865 DOI: 10.1242/bio.059630] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Key Words] [Track Full Text] [Figures] [Journal Information] [Subscribe] [Scholar Register] [Received: 09/12/2022] [Accepted: 11/24/2022] [Indexed: 12/24/2022] Open
Abstract
Biological research is in constant need of new methodological developments to assess organization and functions at various scales ranging from whole organisms to interactions between proteins. One of the main ways to evidence and quantify biological phenomena is imaging. Fluorescence microscopy and label-free microscopy are in particular highly active fields of research due to their compatibility with living samples as well as their versatility. The Imabio Young Scientists Network (YSN) is a group of young scientists (PhD students, postdocs and engineers) who are excited about bioimaging and aim to create a proactive network of researchers with the same interest. YSN is endorsed by the bioimaging network GDR Imabio in France, where the initiative was started in 2019. Since then, we aim to organize the Imabio YSN conference every year to expand the network to other European countries, establish new collaborations and ignite new scientific ideas. From 6-8 July 2022, the YSN including researchers from the domains of life sciences, chemistry, physics and computational sciences met at the Third Imabio YSN Conference 2022 in Lyon to discuss the latest bioimaging technologies and biological discoveries. In this Meeting Review, we describe the essence of the scientific debates, highlight remarkable talks, and focus on the Career Development session, which is unique to the YSN conference, providing a career perspective to young scientists and help to answer all their questions at this career stage. This conference was a truly interdisciplinary reunion of scientists who are eager to push the frontiers of bioimaging in order to understand the complexity of biological systems.
Collapse
Affiliation(s)
- Hana Valenta
- Department of Chemistry, Lab for Nanobiology, KU Leuven, Leuven 3001, Belgium,GDR Imabio, CNRS, Villeurbanne 69100, France,Author for correspondence ()
| | - Nathan Quiblier
- GDR Imabio, CNRS, Villeurbanne 69100, France,Inria, AIstroSight Team, Villeurbanne 69100, France,LIRIS UMR5205, University of Lyon, Villeurbanne, Paris 75015, France
| | - Valerio Laghi
- GDR Imabio, CNRS, Villeurbanne 69100, France,Institut Pasteur, CNRS UMR 3738, Paris, Paris 75005, France
| | - Clément Cabriel
- GDR Imabio, CNRS, Villeurbanne 69100, France,Institut Langevin, ESPCI, Paris, Gif-sur-Yvette 91190, France
| | - Justine Riti
- GDR Imabio, CNRS, Villeurbanne 69100, France,Laboratoire d'Etudes et de Recherches en Immunoanalyse (LERI), CEA, Gif-sur-Yvette, France
| |
Collapse
|