QUASAR--scoring and ranking of sequence-structure alignments

Area of Science:

• Computational biology
• Structural bioinformatics
• Protein structure prediction

Background:

• Sequence-structure alignments are crucial for protein structure prediction methods like fold recognition and homology modeling.
• Existing alignment programs use various criteria (sequence similarity, secondary structure, contact potentials), but selecting the best alignment remains challenging.

Purpose of the Study:

• To introduce QUASAR (quality of sequence-structure alignments ranking), a novel framework for scoring and ranking sequence-structure alignments.
• To provide a unified approach for evaluating combinations of scoring schemes in protein structure prediction.

Main Methods:

• QUASAR offers a framework for combining and scoring sequence-structure alignments using various schemes.
• It enables benchmarking of custom scoring functions against established quality scores (MaxSub, TMScore, Touch, APDB).
• QUASAR includes optimization routines for score combinations based on known structural relationships.

Main Results:

• QUASAR facilitates the identification of well-performing scoring schemes for sequence-structure alignments.
• It allows users to rigorously test novel alignment scoring methods against reliable benchmarks.
• The framework supports optimization of scoring functions for improved protein structure prediction accuracy.

Conclusions:

• QUASAR provides a valuable tool for enhancing protein structure prediction by improving sequence-structure alignment quality assessment.
• The framework supports the development and validation of new scoring strategies in bioinformatics.
• QUASAR aids researchers in selecting optimal alignment scoring combinations for fold recognition and homology modeling.