Fr-TM-align: a new protein structural alignment method based on fragment alignments and the TM-score

Area of Science:

• Structural biology
• Bioinformatics
• Computational biology

Background:

• Protein tertiary structure comparison is crucial in structural biology.
• Existing methods often rely on initial seed alignments refined to optimize structural superposition using metrics like TM-score.
• TM-align, based on TM-score, improved alignment accuracy but had limitations.

Purpose of the Study:

• To develop an improved algorithm for higher quality protein structure alignments.
• To enhance accuracy and coverage in structural comparisons beyond existing methods.

Main Methods:

• Introduced Fr-TM-align, utilizing aligned fragment pairs for initial seed alignments.
• Employed dynamic programming for refining alignments to maximize the TM-score.
• Assessed alignment quality using TM-score and PSI scores, comparing against CE and TM-align.

Main Results:

• Fr-TM-align demonstrated superior structural alignment quality compared to CE and TM-align.
• Achieved approximately 9% higher TM-score and 7% greater coverage on average than TM-align.
• The exhaustive procedure for generating initial seed alignments makes Fr-TM-align computationally more expensive.

Conclusions:

• Fr-TM-align, incorporating fragment alignment and assembly, provides enhanced structural alignments.
• The algorithm offers significant improvements over TM-align for protein structure comparison.
• Source code and executables for Fr-TM-align are publicly available.