A new scoring function and associated statistical significance for structure alignment by CE.

Area of Science:

• Structural bioinformatics
• Computational biology
• Biostatistics

Background:

• Assessing statistical significance in protein structure alignment is crucial for understanding evolutionary relationships and functional similarities.
• Existing methods may lack accuracy or portability across different alignment algorithms.

Purpose of the Study:

• To develop and validate a new scoring function for determining the statistical significance of protein structure alignments.
• To ensure the scoring function is applicable beyond the specific algorithm used for its development.

Main Methods:

• Developed a novel scoring function for protein structure alignment.
• Empirically tested the function using the Combinatorial Extension (CE) algorithm.
• Determined statistical significance (p-values) by curve-fitting score distributions from random protein comparisons (PDB_SELECT and SCOP databases).

Main Results:

• The new scoring function provides statistically robust p-values for protein structure alignment scores.
• The function demonstrated good performance when evaluated using sensitivity, specificity, and Receiver Operating Characteristic (ROC) curves.
• The scoring function proved portable and applicable to other protein structure alignment algorithms.

Conclusions:

• The developed scoring function offers a reliable and versatile tool for assessing the statistical significance of protein structure alignments.
• This advancement aids in more accurate comparative analysis of protein structures in bioinformatics.