The subject of this article was presented as a poster at BEBPA’s 12th Annual Host Cell Protein Conference, from May 14-16, 2024 in College Park, Maryland. You can download the poster here.

Introduction

Managing 2D electrophoresis image data in HCP immunoassay research presents challenges. Traditional naming and organization methods for 2DE or 2D-DIGE experiments often fall short for coverage analysis experiments, like 2D-DIBE.

For example, for quality control purposes, you may want to retain images from all stages of the DIBE generation process, such as gels before and after transfer, and membranes before and after probing, across different fluorophores. You may also need to adjust image acquisition settings such as resolution, PMT voltage, or exposure time to maximize signal detection in the analysis software. This results in a large number of images, especially in studies comparing various antigen and antibody combinations.

Properly organizing, accessing, and analyzing this data is crucial for efficient research.

Common issues with current approaches

Researchers often organize data into multi-level directories categorized by date, experiment type, or other criteria. While this hierarchical structure may seem organized, it becomes cumbersome when trying to find specific images scattered across numerous subfolders. Navigating multiple directories when importing images into analysis software can also be quite tedious.

On top of that, when detailed file names include timestamps, experimental conditions, and other metadata, they often become long and unreadable, making it difficult to identify specific images or select all the data associated with a particular IPG strip. Also, descriptors for experimental conditions can evolve over time. This tends to lead to renaming of files, disrupting filename references in lab notebooks or existing image analyses.

Objectives of our approach

We aim to streamline the management of image files and metadata from 2D electrophoresis experiments, focusing on specific objectives:

Workflow integration

The approach is tailored to support essential workflow steps, allowing researchers to establish file names as they complete experiment details in the lab notebook—even before image capture. It therefore avoids using image acquisition timestamps in file names, in favor of a more predictable and structured naming system.

Metadata should be recorded immediately after image acquisition, while details of the experiment and acquisition settings are still fresh. This ensures that the metadata accompanies the image files into the analysis software, optimizing setup and enhancing analysis and interpretation.

Streamline consistency and referencing across platforms

By applying consistent file names that persist over time and across platforms (image acquisition devices, lab notebooks, storage systems, analysis software), we ensure that files can always be traced back to their entries in lab notebooks or file systems. While consistent file names alone don’t ensure images haven’t been modified (digital fingerprints are required for that), they help team members verify they are referencing the correct data.

Safeguard data lineage and integrity

Our system is designed to track data lineage, making it easy to locate all images derived from the same IPG strip. This facilitates rigorous quality control checks and troubleshooting, and contributes to robust experimental validation. To prevent any confusion or misinterpretations during the analysis phase, our system explicitly identifies the nature and stage of each image—whether it pertains to a gel before or after transfer, or a blot before or after probing. This ensures that every team member can accurately interpret the data, maintaining the integrity of the research process and results.

Refined strategies for image data management

To overcome the limitations of traditional strategies, we concentrate on several key aspects.

Flat folder structure

We recommend a single folder per major experiment or project to facilitate quick access and effective software import. If images from the same acquisition (e.g., different dye images of the same DIGE gel or DIBE blot) are saved in a subfolder, they may remain if referenced in a top-level (.ds) file.

File naming policy

When avoiding hierarchical folders, a robust naming convention becomes essential. It should ensure:

  • Consistency – Names remain constant across all platforms.
  • Uniqueness – Prevents conflicts and supports single-folder storage.
  • Traceability – Enables tracking of all images from an IPG strip.
  • Brevity – Simplifies reading and retrieval; avoids issues with long paths.
  • Predictability – Allows names to be set during planning, independent of acquisition time.
  • Objectivity – Team members follow a simple, predefined policy without subjective decisions.

We recommend the following naming convention elements:

e.g. 5974Gat Cy3 A01(2)

Strip number Use the last digits of the IPG strip number, consistently from the same batch for a specific experiment.
Object type Indicate object and status (e.g.  G – gel before transfer, Gat – gel after transfer, M – membrane before probing, B – membrane after probing).
Stain/label Specify the stain or label used, such as Cy2, Cy3, or Cy5. This detail is crucial for DIGE/DIBE projects, allowing Melanie to automatically group images into DIGE or DIBE objects. Note that .ds files do not contain dye labels, but the referenced images within them do.
Acquisition code Optionally add a code for image acquisition parameters, kept in the Metadata File. This code is identical for images of a given DIBE or DIGE object (e.g. A01 = 50 µm resolution, PMT Cy3 490V, PMT Cy5 420V). For duplicate acquisitions with the same settings, add a replicate number.

For instance, the filename ‘5974Gat Cy3 A01(2)’ refers to an image from the Cy3 channel of a gel associated with the IPG strip ending in 5974, after protein transfer to the membrane, using the initial set of acquisition parameters. The ‘(2)’ indicates this object was previously captured under the same conditions, perhaps on a different date or scanning area. This image typically serves as a control to verify successful protein transfer to the membrane. No spots should be visible on it.

By organizing images by filename, you can effortlessly locate and select all other images associated with IPG strip 5974.

Use of aliases in image analysis

We have stressed that file names like ‘5974Gat Cy3 A01(2)’ should remain fixed to ensure consistency across platforms. However, these names often lack detail for effective image analysis, where more descriptive names like ‘CHO-HCP LMW’ or ‘anti-CHO HCP Ab’ prove useful, providing information about antigens, antibodies, or other parameters.

Responding to this need, the upcoming release of Melanie will introduce aliases, or alternative image names, in all image labels and reports. These aliases will coexist with the unchanged original file names and can be customized to suit various needs—from detailed names for internal use to anonymized versions for publication.

You can either manually edit aliases or import them from the Metadata file. There you can automatically create them based on existing data fields.

Using metadata effectively

Instead of embedding all data within filenames, we store detailed metadata in an Excel file located in the same directory as the image files. This file always accompanies the image data.

The Excel Metadata File logs detailed experiment information, including columns for file name, alias, acquisition date and time, detailed acquisition parameters, sample information, and experimental notes. Excel allows easy data management through features like copying, using formulas, and data validation tools (e.g., restricting data types or values with dropdown lists). These functionalities enhance metadata quality and allow for efficient sorting, filtering, searching, and retrieval.

Additional sheets in the Excel Metadata File can include picklists, a data dictionary, and data specifications.

This metadata serves multiple purposes:

  • Helps quickly locate images that meet specific criteria.
  • Automatically generate file names or aliases through concatenation.
  • Enables direct import of aliases and metadata into Melanie software (in the upcoming release).
  • Allows copying of metadata subsets into your lab notebook.
  • Facilitates creation of anonymized metadata files for sharing with collaborators.

Conclusion

Managing 2D electrophoresis data requires thoughtful strategies that ensure accessibility and detailed traceability. Adopting a flat folder structure, a precise file naming policy, and a robust Excel-based metadata system allows research teams to enhance workflow efficiency and ensure data consistency.

Specific benefits of this approach include:

  • Safeguarding data lineage and integrity.
  • Streamlining consistency and referencing across platforms.
  • Reducing file retrieval times.
  • Facilitating the import and reuse of metadata.
  • Minimizing labeling and analysis errors.
  • Optimizing image analysis setup and results interpretation.
  • Simplifying auditing procedures.

While tailored for 2DE coverage experiments, these principles are also applicable to conventional protein expression analysis.

Your move: Implementing these solutions

We encourage you to tailor these strategies to your research requirements. To assist you, we offer a downloadable Excel Metadata File template. We welcome your feedback on the document and invite you to share any additional tips or alternative methods you may have discovered in your practice.

Our aim is to provide a foundational, effective approach to managing 2D gel and blot data. For further guidelines on data organization and metadata usage, you can refer to the articles below.

  1. Monash University Library. (n.d.). Organising data. Monash University. Retrieved May 12, 2024, from https://www.monash.edu/library/researchers/data-collection-management/guidelines/organising-data
  2. University of Wisconsin-Madison. (n.d.). Data literacy. Retrieved May 12, 2024, from https://data.wisc.edu/data-literacy/document/
  3. DataONE. (n.d.). Best practices for data description. Retrieved May 12, 2024, from https://dataoneorg.github.io/Education/bp_step/describe/
  4. Research Data Services, UCSB Library. (n.d.). Documenting and organizing data. University of California, Santa Barbara. Retrieved May 12, 2024, from https://rcd.ucsb.edu/resources/data-resources/documenting-organizing