PXMeter

PXMeter supports consistent and reproducible evaluation of diverse predictive models for biomolecular complex structures.

Key Features:

• Unified Benchmarking Framework: PXMeter provides a unified and reproducible benchmarking framework for evaluating structure prediction models.
• Benchmark Dataset: PXMeter includes a high-quality benchmark dataset curated from recently deposited structures in the Protein Data Bank (PDB), with entries manually reviewed to exclude non-biological interactions.
• Model Evaluation: PXMeter facilitates assessment of predictive models across various biological complex structures including protein-protein and protein-RNA interactions.

Scientific Applications:

• Structure Prediction Assessment: PXMeter enables comprehensive benchmarking of structure prediction models such as Chai-1, Boltz-1, and Protenix.
• Biomolecular Complex Analysis: PXMeter supports evaluation of models predicting protein-protein and protein-RNA interactions.

Methodology:

PXMeter curates benchmark datasets from recently deposited PDB structures with manual review to exclude non-biological interactions, then applies a unified evaluation framework to assess predictive models.