Imaging

Point people/organizers: Lacra Bintu and Caterina Strambio De Castillia

4DN Google Drive folder: https://drive.google.com/drive/folders/1EmBUDLIU0WDf-d09xB8YLBhidf_2A7c6?usp=sharing

4DN Calendar: https://wiki.4dnucleome.org/4dn:calendar_4dn (click here to add this specific WG calendar to your personal calendar)

Agenda/Minutes: https://docs.google.com/document/d/1cFqYZvOTPmLGUUwxohFchCD7G3v0qpkfN_HC33KX7Gc/edit?usp=sharing

Meeting attendee spreadsheet: https://docs.google.com/spreadsheets/d/1WeR515p6DT3RXG9FTzIDNtSjRetHbrbwJxmBo_9FLTc/edit?usp=sharing

Email list: image@4dnucleome.org

Slack channel: #wg-imaging

One of the main aims of 4DN Phase 2 is to improve the functional understanding of chromatin organization in health and disease. To this aim, it will be very important to define shared methods for the integration of genomics with imaging data. In this context the Imaging Working Group will serve two functions:

1) As a forum with the aim of defining clear short, mid-, and long-term strategies and solutions to improve the documentation, quality, reproducibility, sharing, and reusability value of imaging data within the 4DN and broader community. This will include the close collaboration with the Integrating Imaging and Omics Working Group and the exchange of information and collaboration with other NIH consortia and global imaging initiatives active in this space (e.g., BINA, Global Bioimaging, QUAREP-LiMi, etc.).

2) As a coordination platform to initiate sub-groups (SWG) that focuses on concrete barriers related to imaging within 4DN, as a discussion forum about solutions and as a place for such SWG to report the outcome of their work before implementing policies via the DCIC. Subgroups can have any format, should be led by their initiators, and typically aim to solve problems (or define clear strategies for the resolution of more complex problems) on a short time scale of 3-6 months. The IWG encourages all 4DN members to initiate SWGs based on their interests and needs. SWG should explicitly be open to being initiated by postdocs and staff scientists.

Work from IWG has been featured in a FOCUS issue on Nature Methods in December 2021

• Reporting and Reproducibility in Microscopy –> FOCUS ISSUE

• Community-driven 4DN-BINA-OME Microscopy Metadata Specifications

• Micro-Meta App tool for collecting Microscopy Metadata (now in use on the 4DN Data Portal)

The purpose of this work is to develop a standard format to allow the exchange of imaging results and encourage the development of shared analysis, modeling and visualization pipelines.Currently, we are working on two formats to exchange FISH Omics results.

1. FISH omics Format for Chromatin Tracing (FOF-CT) is released and provisionally approved by the SC (see below)
2. FISH omics Format for Genomic Single Molecule Localization is in progress and current status can be found at the bottom of the exchange data format subgroup page.

The first output of this sub-group is the proposed FISH Omics Format for Chromating Tracing (FOF-CT). We encourage the community to provide feedback ahead to approval from the November Steering Committee meeting. The format is available on our ReadTheDoc page and on GitHub:

• Please refer to theReadTheDocs page for any and all information about this format and to download example files to get started.
• Please report any questions, issues, and proposals for change on theGitHub issue page
• Please refer to this example dataset that was deposited on the 4DN-Data Portalusing FOF-CT: * Liu M, Lu Y, et al.,Nature communications2020 →https://data.4dnucleome.org/experiment-set-replicates/4DNESXYFBZMI/

IMPORTANT: All previous GoogleDoc and GoogleSheet representations of FOF-CT are now deprecated and no longer maintained.

1. SWG Image Data Documentation and Submission: Lead: Andrea Cosolo, Andrew Schroeder, and Caterina Strambio De Castillia (alphabetic order). Mission Statement:The aim of this SWG is to improve awareness of 4DN members of the tools and procedures made available by the DCIC for image data documentation, hold appropriate training sessions and encourage rapid and stringent data submission.
2. SWG ImagingOmics Exchange Data Formats for Genomic Imaging Results . Lead: Alistair Boettiger, Steve Wang (alphabetic order). Mission statement: This sub-group will develop recommended data formats and metadata specifications for sharing 3D chromatin folding traces obtained from multiplexed DNA FISH and other in-situ imaging-based omics approaches, with the flexibility to include additional multi-omic imaging data to be mapped to these structural data, such as RNA expression, protein distribution, cell morphology, or tissue coordinates.

1. SWG Community Outreach and Interaction on Imaging Data Documentation, QC, Specifications, and Standards Lead: Caterina Strambio De Castillia. Proposed Mission Statement: Work of the 4DN Phase 1 IWG has catalyzed activity in the bioimaging community at multiple levels from individual researchers to imaging facilities to infrastructure stakeholders. This working group aims to conduct work that will lead to Image Data Standards (DIN/ISO, NIST, IEEE, etc.) by coordinating work conducted within 4DN with work being conducted by the community at large. The aim of this work is to maximize the quality, reproducibility of 4DN datasets and tools, maximize their impact on the community, and support the community in further developing tools and specifications.
2. SWG Pilot Image Data Sharing Project. Lead: to be decided. Proposed Mission Statement:This sub-group will work in coordinationwith theIntegrating Imaging and Omics working group to identify and enroll 4DN member laboratories to participate in pilot projects aimed at evaluating the scientific value of creating a 4DN-wide repository for deposition and exchange of richly documented raw image data
3. SWG NIH consortia coordination. Lead: to be decided. Proposed Mission Statement:With several other NIH common fund consortia tackling similar problems this sub-group focuses on coordinating work with other consortia including exchanging information and attending relevant working groups in other consortia.

Next meeting: 3/2/2021

The work of the Imaging Working Group is closely connected with the work of the following:

• Integrating Imaging and Omics Working Group
• Real-time Chromatin Dynamics and Function Interest Group

Ensuring Image Quality through the use of shared Quality Control procedures is essential to reproduce, interpret and compare microscopy data especially when image analysis (i.e., spot localization and tracking) is needed for results interpretation. As a starting point for this effort, during Ph1 the ISWG invested close to a year to create a common biological sample ( Szymborska et al., 2013) that could be used to compare results across labs. Scientists at EMBL, Cornell, the Hutch, and several other sites were engaged in acquiring images using the same biological sample to assess to what extent images and image metadata could be compared assuming a basic knowledge of the microscope. As expected by several of the ISWG leads, this effort failed dramatically. Specifically, even though all laboratories acquired images of a shared benchmarking sample, resulting images displayed significant variations across sites preventing direct comparison. In addition, the metadata collected at individual acquisition sites also varied dramatically resulting in widely different information content. And while a similar sample was used for comparing Single-molecule localization microscopes (SMLM; Thevathasan et al., 2019), tests conducted within 4DN laboratories revealed significant obstacles for using this biological sample across different microscope modalities. As a result, it was decided that the 4DN-Phase 1 ISWG should instead develop the following:

• The current version of the 4DN-BINA Microscopy Metadata Guidelines is available on GitHub and published on ArXiv.org

While the power of digital light microscopy is undeniable, rigorous record-keeping and quality control are required to define the limits to which the results of imaging experiments may be interpreted (quality), reproduced (reproducibility), used to extract reliable information and scientific knowledge, and shared for further analysis (value). At the same time, it is also clear that different experiments have different reporting requirements to ensure quality, reproducibility, and value. To address this issue, the 4DN-Phase 1 the ISWG working in conjunction with the Bioimaging North America (BINA) Quality Control and Data Management Working Group (QC-DM-WG) has developed a flexible and future proof framework that extends the specifications implemented in the Open Microscopy Environment (OME) Image Data Model (https://genomebiology.biomedcentral.com/articles/10.1186/gb-2005-6-5-r47; https://rupress.org/jcb/article/189/5/777/35828/Metadata-matters-access-to-image-data-in-the-real). This approach has been endorsed by several initiatives around the world including BINA, Quality Assessment, and Reproducibility for Light Microscopy (QUAREP-LiMi; https://arxiv.org/abs/2101.09153), Global Bioimaging (GBI), the Allen Institute of Cell Science and OME. It is fair to say that 4DN has launched a much-needed process that forms the basis of a wide community effort involving all major stakeholders (importantly including microscope manufacturers) and that has the potential to result in the first-ever widely accepted community-driven Microscopy Metadata specifications that scale with the complexity and intent of the experimental design, instrumentation and analytical requirements.The Tiered OME-extension system initially proposed by 4DN and revised by the BINA-QC-DM-WG provides the flexibility to:

• Consider the intent and complexity of an experiment (think: counting nuclei stained with DAPI vs. tracking single molecules in living cells with ms and nm resolution in 20 colors),
• Leverage the existing OME image data specifications (i.e., the OME Data Model forms the basis of Bio-Formats, which is the library that allows all existing image data formats to be read by different software tools including but not limited to Fiji).
• Create reporting blueprints for specific assays so that the data can be reproduced, shared, and compared by multiple labs
• Create general reporting blueprints that can be adopted by different communities while allowing easy mapping of variables.

During 4DN Phase 1, in response to the need to facilitate the rigorous collection of microscopy metadata, the Strambio De Castillia lab in collaboration with the Grunwald lab and the DCIC has developed an interactive software tool called the Micro-Meta App. Micro-Meta App can run both locally and within the DCIC webpage. It reads metadata that is stored within an image file and automatically fills the corresponding fields in a Tier-aware manner. The App displays a graphical rendition of a microscope and allows to interactively add individual parts while allowing the user to fill in metadata values that are needed at the chosen Tier level and providing feedback on missing information. If you do not have that information at hand the app provides you with the knowledge to ask your sales rep to please provide it to you.

MicroMeta provides a user interface to collect metadata about microscopes and, soon, their settings based on the 4DN-BINA-OME specifications.

• The MicroMeta App is integrated into the 4DN Data Portal allowing users to create, edit, save, and clone microscopes.
• To learn how to use Micro-Meta App please see the Micro-Meta App documentation
• The source code for the current version of the Micro-Meta App is available on GitHub

• Optical calibration standards to ensure the quantitation and comparison of the resolving power and chromatic discrimination capacity of individual microscopes. Different optical standards were evaluated, and multicolor beads were considered the best option while retaining affordability (see attachments).
• Standard and tools for Excitation Intensity and Emission Detection calibration. To truly compare image data, it is necessary to “translate” intensity (i.e., grey values), which is measured in arbitrary units into photon counts. This allows us to compare raw data and describe the extent to which high-precision localization information can be extracted using advanced methods. While this can be done today using standard tools (i.e., power meter) it was reasoned that it was necessary to develop an easy-to-use tool to enhance comparability. As a result, the Zipfel and Grunwald labs developed “best practice calibration” cheat sheets (see attachments) for wide-field and confocal microscopes. Based on the detector calibration light source used by the Warren lab the Grunwald lab developed and started producing a combined power sensor, light source calibration sensor tool called MetaMax. Detector calibration works for point detectors like PMTs, APD, SPADs, and similar as well as for cameras (including CCDs, EMCCDs, and sCMOS). Using detector calibration, the grey values in an image can be back-translated into photons, which opens a number of ways to compare raw data information content. Detector calibration is a standard procedure in many physics and engineering disciplines. While detector calibration can be done manually a tool as MetaMax aims at making this process easy and fast for non-experts. MetaMax can be attached to the microscope in place of an objective, and it is ready to use out of the box. The 5th generation prototype is currently being calibrated to be tested at the Allan Institute of Cell Science and has a waitlist of some tens of BINA and QUAREP-LiMi member labs eager to adopt it.

To exchange quantitative results from imaging experiments, the 4DN-Ph1 IWG developed three data exchange formats that are currently utilized by the DCIC and could be applicable for Ph2:

1. DNA FISH Spot Localization _v01 (Draft),
2. Segmentation _v01 (Draft), and
3. Single Particle Tracking (Approved by Phase I IWG)

These data exchange standards are extremely relevant for Phase 2 and could serve as a great starting point for further development. In addition, it is important to note that as part of the effort to develop a Next-Generation File Format for image data, the OME consortium is developing standards for recording Region of Interests (ROI), spots, and trajectories which is extremely relevant to this topic.