MethodsJ2: A Software Tool to Capture Metadata and Generate Comprehensive Microscopy Methods Text

Proper reporting of metadata is essential to reproduce microscopy experiments, interpret results and share images^1,2. The lack of methods reporting in microscopy is evident in that few research articles pass a test for the minimal information required to reproduce experiments¹ (~17% of 240 articles with 1,500 figures with images). The problem is compounded by the number and variety of microscope modalities, options and associated components. Automation has distanced researchers from the technical parameters so it is difficult for them to know what information needs to be reported. MethodsJ2 is an ImageJ/Fiji based software tool that aims to improve reproducibility in microscopy.

To properly evaluate and reproduce microscopy images, information about sample preparation, experimental conditions, microscope hardware, image acquisition settings and image analysis parameters is required. This information is called “metadata” and is defined as “a set of data that describes and gives information about other data”. Researchers involved in the 4D Nucleome initiative³ and Bioimaging North America (BINA) (https://www.bioimagingna.org/) have developed extensive community driven Microscopy Metadata specifications^4,5. These specifications build on a previous Open Microscopy Environment (OME) model⁶ and include an in-depth community driven Microscopy Metadata model for light microscopy termed “4DN-BINA-OME”⁴. The model scales with experimental design, instrument complexity and the degree to which image processing and quantitative image analysis is required for interpreting results. This ensures essential information is included while the burden on experimental scientists to collect and report metadata is minimized⁷.

Microscope Metadata guidelines^8–10, examples of what can go wrong if metadata is not reported¹¹ and the importance of measuring and reporting microscope quality control¹² have been published. Increased awareness/education around Microscopy Metadata and straightforward accessible tools are vital for successful implementation. MethodsJ2 is an extensible open-source microscopy methods reporting software tool that runs in ImageJ/Fiji and builds on MethodsJ^1,13,14. It captures Image Metadata from multiple sources, consolidates it and automatically generates methods text for publication. Integration with ImageJ/Fiji should make it broadly available to experimental scientists.

MethodsJ2 automatically gathers metadata from the image using OME BioFormats (e.g. pixel size, magnification) and captures Microscopy Metadata from a Microscope.JSON file generated using Micro-Meta App^5,15. Micro-Meta App is a companion software tool that guides researchers step-by-step in the collection of community standardized Microscopy Metadata for a specific microscope⁴. MethodsJ2 also guides the user to enter specific Experimental/Sample Metadata (e.g. cell type, dyes). Finally, the software guides the user through a step-by-step validation of the metadata. To improve tracking of imaging facility impact, acknowledgement text, including a facility Research Resource ID (RRID, https://scicrunch.org/resources) can be added to the script. The methods text is then automatically generated but must be reviewed and edited.

Detailed MethodsJ2 workflow (Figure 1).

Figure 1: MethodsJ2 Workflow Overview.

Steps required to automatically generate microscopy methods text. Image metadata is collected from the manufacturer metadata in the image file using the OME TIFF tools. Hardware metadata is collected from a Micro-Meta App Microscope.JSON file. It is recommended to have an experienced microscopist or imaging scientist guide researchers through the methods text generation and validation process. Acknowledgement text can be added to the script by imaging scientists in the microscopy platform including a RRID.

Supplemental materials provide a more detailed workflow and sample Microscope Metadata.

1. Use Micro-Meta App to create and save a Microscope.JSON file. Note: Give components detailed names as this text populates the methods text. Put “63x/1.4 NA Plan-Apochromatic oil immersion” not “63x”.

2. Download the MethodsJ2 script (file named: MethodsJ2_v1_2_.py), an example Microscope.JSON file and an example image file from GitHub (https://github.com/ABIF-McGill/MethodsJ2). Download and install ImageJ/Fiji (https://fiji.sc/).

3. Drag the MethodsJ2 script file and drop it on the ImageJ/Fiji toolbar. The Script Editor will open, then press “Run”.

4. Select an image file. The Image Metadata is automatically extracted. Sample information can be added manually. Select a Microscope.JSON file for the corresponding microscope.

5. The user is guided step-by-step to validate the metadata and input critical hardware and settings information. Note: Have an experienced microscope user or imaging scientist help with this step.

6. Click “OK”. Draft text is automatically generated and appears in a popup window, is copied to the clipboard and can be pasted into a manuscript. A .csv file of the Microscope Metadata is generated and saved. See the sample .csv file included as supplemental material and on the GitHub portal. Note: It is the responsibility of the experimental scientists to review the draft text to ensure it is accurate.

Comprehensive methods reporting is essential for reporting imaging data, sharing images and emerging new methods^16–22. Progress along the path of rigor and reproducibility is essential for high quality microscope-based science and is a shared responsibility. Experimental scientists must use due diligence to understand the fundamentals of the technologies and required Microscope Metadata their research relies on. Imaging scientists need to educate experimental scientists, so they understand what metadata needs to be reported and why. Microscope manufacturers ought to integrate, automate and report Microscope Metadata. Scientific publishers and reviewers have a duty to promote community-based guidelines^4,6,23 and ensure published microscope images meets a minimum standard. Funding agencies need to uphold high quality reproducible microscope images and ensure detailed Microscopy Metadata is available when images are publicly shared.

MethodsJ2 and two companion software tools, Micro-Meta App¹⁵ and OMERO.mde²³, advance rigor and reproducibility in microscopy (Supplemental Figure), but there are still challenges. Microscope Metadata is often limited, not in standard formats, not accessible due to proprietary microscope manufacturer software and/or lost when images are saved and opened with third-party software⁴. Microscope manufacturers need to work with the global community through organizations like Quality Assessment and Reproducibility for Instruments & Images in Light Microscopy (QUAREP-LiMi)^24,25 to automate metadata collection, ensure it conforms to community standards^4,6,23 and make it readily available. Implementation and evolution of MethodsJ2, Micro-Meta App¹⁵ and OMERO.mde²³, will advance rigor and reproducibility in microscopy, promote transparency and reproducibility and help stakeholders ensure Microscopy Metadata is documented and reported.

DATA AND CODE AVAILABILITY STATEMENT:

Data in the form of a sample image and MICROSCOPE.JSON file are available at https://github.com/ABIF-McGill/MethodsJ2. Full source code and step-by-step instructions are available at https://github.com/ABIF-McGill/MethodsJ2 and 10.5281/zenodo.5172827.