BDM

BDM evaluates protein complex structure models using a distance difference matrix to quantify differences between predicted and reference structures.

Key Features:

• Distance Difference Matrix (BDM): Compares pairwise inter-residue distances between predicted and reference protein complex structures to produce a BDM score.
• Correlation with RSMD: Establishes a linear correlation between BDM and Root Mean Square Deviation (RSMD).
• No structure alignment required: Eliminates the need for structure alignment by addressing receptor-ligand differentiation.
• Benchmarking on CASP datasets: Demonstrated performance on CASP14 and CASP15 test sets.
• Addresses limitations of existing CASP metrics: Targets shortcomings of methods such as DockQ used in CASP evaluations.

Scientific Applications:

• Evaluation of predicted protein complex structures: Provides a quantitative metric for assessing accuracy of protein complex models.
• Benchmarking and method comparison: Enables comparison of scoring methods and assessment protocols using CASP14 and CASP15 test sets.
• Receptor–ligand complex assessment: Facilitates evaluation of receptor-ligand complexes without requiring receptor/ligand differentiation via alignment.

Methodology:

Compute a distance difference matrix between predicted and reference complex structures to derive the BDM score, compare BDM values to Root Mean Square Deviation (RSMD), and evaluate performance on CASP14 and CASP15 test sets.