Mass Dynamics 1.0: A streamlined, web-based environment for analyzing, sharing and integrating Label-Free Data

Joseph Bloom, Aaron Triantafyllidis, Paula Burton, Giuseppe Infusini, Andrew I. WebbbioRxiv 2021 · DOI

TL;DR

Mass Dynamics 1.0 (MD 1.0) is a new web-based platform that streamlines the analysis, sharing, and integration of label-free quantitative (LFQ) proteomics data. It offers an automated, user-friendly environment that aims to improve reproducibility and collaboration in proteomics research, while delivering results comparable to established tools like Perseus.

What Problem Does This Paper Address?

Proteomics, the study of proteins, is crucial for understanding biological processes and disease. A key technique is Label-Free Quantification (LFQ), which measures protein abundance across different samples without using isotopic labels. While powerful, LFQ analysis involves complex pipelines and numerous software tools, often leading to challenges in reproducibility, data sharing, and ease of use, especially for researchers who aren't bioinformaticians. Existing tools, while capable, can be overwhelming due to their extensive functionality, making straightforward analysis time-consuming. This paper addresses the need for a more integrated, automated, and collaborative platform that simplifies LFQ data analysis while maintaining high reliability.

Methods

MD 1.0 is built on a modern software stack (JavaScript and Rails for the web component) and uses R for its processing engine, hosted on Amazon Web Services (AWS). This architecture ensures repeatable processing with identical software, parameters, and code. Key features include:

  • Automated Analysis: MD 1.0 takes MaxQuant output files and automates the entire LFQ analysis pipeline, minimizing user parameter choices.
  • Cloud-Based Architecture: Enables secure storage, sharing, and collaboration on experimental data.
  • Quality Control (QC) Reports: Automatically generates comprehensive QC reports to assess experiment integrity.
  • Over-Representation Analysis (ORA): Integrates with the Reactome API to identify enriched biological pathways within user-defined protein lists.
  • User Interface: Designed for ease of use, with distinct tabs for analysis, QC reports, insights, results files, export, and experiment design.

Datasets

The authors evaluated MD 1.0 using several established LFQ benchmarking datasets:

  • iPRG2015 dataset: Contains raw data for S. cerevisiae samples spiked with known proteins, providing a "ground truth" for evaluation.
  • Dynamic Range Dataset (PXD000279): Uses "Universal Protein Standards" to test LFQ accuracy over a wide dynamic range.
  • HER2 Dataset (PXD002057): Compares protein abundance between a parental cancer cell line and a resistant derivative, representing a more "real-world" scenario.

These datasets were chosen to verify that MD 1.0 recovers results similar to Perseus, a widely used analytical tool.

Key Insights and Ideas

  • Bridging the Gap: MD 1.0 aims to democratize proteomics analysis by providing a user-friendly interface that hides much of the underlying complexity, making advanced analysis accessible to a broader range of users.
  • Reproducibility through Automation: By automating the analysis pipeline and using a consistent cloud-based environment, MD 1.0 significantly enhances the reproducibility of LFQ results.
  • Collaboration is Key: The platform's integrated sharing features are designed to foster collaboration, allowing researchers to easily share their findings and receive assistance.
  • Beyond Basic Analysis: The inclusion of ORA with Reactome integration adds a layer of biological interpretation directly within the platform, moving beyond just quantitative results.

Key Results

  • High Accuracy: MD 1.0 reproduced results comparable to Perseus across multiple benchmarking datasets, with Pearson correlations for log fold changes consistently above 0.9.
  • Reliable Differential Expression: Binary evaluation of differential expression predictions showed high concordance between MD 1.0 and Perseus, with very few false negatives or positives.
  • Comprehensive QC: The automated QC reports provide users with confidence in the quality and integrity of their experimental data.
  • Ease of Use: The platform requires minimal parameter choices from the user, simplifying the analysis process.

Limitations and Future Work

  • MaxQuant Dependency: Currently, MD 1.0 primarily accepts output from MaxQuant. The authors suggest that the platform is designed to be adaptable to data from other preprocessing tools in the future.
  • Enrichment Analysis Expansion: While ORA is integrated, the authors propose expanding analysis capabilities to include other enrichment methods like Gene Set Enrichment Analysis (GSEA) and integrating with more third-party databases (e.g., GO, DAVID, KEGG, Drugbank).
  • User Interface Improvements: Further enhancements to the user interface are envisioned, particularly in annotation and sharing utilities.
  • Leveraging Cloud Infrastructure: The authors suggest exploring how the cloud-storage element can be used for cross-referencing experiment data and insights.

Summary updated

Abstract

Abstract Label Free Quantification (LFQ) of shotgun proteomics data is a popular and robust method for the characterization of relative protein abundance between samples. Many analytical pipelines exist for the automation of this analysis and some tools exist for the subsequent representation and inspection of the results of these pipelines. Mass Dynamics 1.0 (MD 1.0) is a web based analysis environment that can analyze and visualize LFQ data produced by software such as Maxquant. Unlike other tools, MD 1.0 utilizes cloud-based architecture to enable researchers to store their data, enabling researchers to not only automatically process and visualize their LFQ data but annotate and share their findings with collaborators and, if chosen, to easily publish results to the community. With a view toward increased reproducibility and standardisation in proteomics data analysis and streamlining collaboration between researchers, MD 1.0 requires minimal parameter choices and automatically generates quality control reports to verify experiment integrity. Here, we demonstrate that MD 1.0 provides reliable results for protein expression quantification, emulating Perseus on benchmark datasets over a wide dynamic range. The MD 1.0 platform is available globally via: https://app.massdynamics.com/ . Contact [email protected]
BibTeX
@misc{bloom2021mass,
  author = {Joseph Bloom and Aaron Triantafyllidis and Paula Burton and Giuseppe Infusini and Andrew I. Webb},
  title = {{Mass Dynamics 1.0: A streamlined, web-based environment for analyzing, sharing and integrating Label-Free Data}},
  year = {2021},
  doi = {10.1101/2021.03.03.433806},
  howpublished = {bioRxiv},
}
Figure