SARST2 high-throughput and resource-efficient protein structure alignment against massive databases.

Area of Science:

• Computational Biology
• Bioinformatics
• Structural Biology

Background:

• The rapid growth of protein structural data presents significant computational challenges for researchers.
• Existing protein structural alignment tools may struggle with the increasing scale and complexity of biological Big Data.

Purpose of the Study:

• To develop an efficient and accurate protein structural alignment search algorithm to address growing computational demands.
• To present SARST2, a novel algorithm designed to assist researchers in analyzing large protein structure datasets.

Main Methods:

• SARST2 integrates primary, secondary, and tertiary protein structural features with evolutionary statistics.
• It utilizes a machine learning-enhanced filter-and-refine strategy, diagonal shortcut for word-matching, and a weighted contact number-based scoring scheme.
• A variable gap penalty based on substitution entropy is incorporated for enhanced alignment accuracy.

Main Results:

• SARST2 demonstrates superior accuracy compared to state-of-the-art methods in large-scale benchmarks.
• It achieves significantly faster search speeds and requires substantially less memory than BLAST and Foldseek for AlphaFold Database searches.
• The algorithm enables massive database searches on ordinary personal computers.

Conclusions:

• SARST2 provides an efficient and powerful analysis engine for protein structural Big Data.
• Its performance advancements in speed and memory usage facilitate broader accessibility for biological research.
• The algorithm supports researchers in pushing the frontiers of biological sciences and technology through advanced computational analysis.