Nature's economy as blueprint: regional overlap & block-based algorithm for protein design

Area of Science:

• Computational biology
• Structural biology
• Bioinformatics

Background:

• Nature utilizes modular recombination for protein design, optimizing adaptation and minimizing energetic costs.
• Identifying conserved structural scaffolds is crucial for understanding protein function and evolution.

Purpose of the Study:

• To present ROB-Fold, a deterministic rule-based algorithm for identifying conserved structural scaffolds across protein families.
• To emulate natural evolution for discovering biologically meaningful patterns in protein structures.

Main Methods:

• ROB-Fold analyzes experimentally validated and AI-predicted protein models (AlphaFold, Protein Data Bank).
• It employs motif-centered analyses, bidirectional expansions, Chou-Fasman structural equivalence, and Root Mean Square Deviation (RMSD) for quantification.
• The algorithm detects recurrent sequence and structural elements anchored on biologically significant patterns.

Main Results:

• ROB-Fold identified a conserved thioredoxin-like scaffold in protein disulfide isomerases (PDIs) centered on the CXXC motif.
• It revealed a recurrent pattern in HSP70 chaperones defining the nucleotide-binding domain, associated with the EEVD motif.
• Conserved regions across both families showed RMSD values below 15 Å, confirming structural persistence.

Conclusions:

• Explicit structural rules can identify evolutionary invariants, offering an interpretable and efficient alternative to stochastic models.
• ROB-Fold provides a transparent, biologically grounded framework for scaffold identification.
• The algorithm serves as a guide for the rational design of novel proteins.