QSProteome: a community-driven interactive platform for large-scale exploration and evaluation of predicted protein...

Area of Science:

• Structural biology
• Computational biology
• Bioinformatics

Background:

• Quaternary protein structure modeling and validation are crucial for understanding biological functions.
• Existing resources often lack comprehensive, community-driven refinement capabilities.
• Scalable platforms are needed to manage and analyze large-scale proteome structural data.

Purpose of the Study:

• To introduce QSProteome, a community-scale platform for modeling, evaluating, and refining quaternary protein structure.
• To provide a dynamic, scalable infrastructure for proteome-scale structural biology.
• To enable community-led assessment and refinement of protein complex structures.

Main Methods:

• Hosting over 35,000 modeled protein assemblies from BioCyc and ComplexPortal.
• Utilizing interactive web pages with 3D visualization, confidence metrics, and functional annotations.
• Implementing a cloud-based server for automated processing and a gamified re-curation workflow.
• Enabling community submission and validation of protein models.

Main Results:

• QSProteome hosts 35,528 unique modeled assemblies covering extensive gene sets.
• The platform processed and validated over 54,000 models within 14 weeks.
• A community-led curation workflow assessed 1547 ABC transporter complexes.
• Interactive features facilitate detailed analysis and validation of each model.

Conclusions:

• QSProteome provides a unified, searchable, and extensible framework for proteome-scale structural biology.
• The platform enhances accessibility and reusability of structural data.
• It empowers collaborative refinement and discovery within the structural biology community.