Detection of conserved segments in proteins: iterative scanning of sequence databases with alignment blocks

Area of Science:

• Bioinformatics
• Computational Biology
• Proteomics

Background:

• Analyzing large protein sequence databases is crucial for understanding protein function and evolution.
• Identifying conserved segments within protein families is essential for motif discovery.
• Existing methods may have limitations in detecting novel or subtle conserved regions.

Purpose of the Study:

• To develop and validate an iterative computational approach for identifying conserved segments in protein sequence databases.
• To establish a robust method for generating position-dependent weight matrices from coevolving segments.
• To detect novel conserved motifs with potential biological importance.

Main Methods:

• Iterative database scanning using an evolving position-dependent weight matrix.
• Construction of weight matrices from aligned conserved segments.
• Utilizing a logarithm-of-odds, Bayesian-based approach with Dirichlet mixture priors for matrix calculation.
• Setting iteration cutoffs based on expected random model score distributions.

Main Results:

• The iterative procedure converged for all alignment blocks studied, requiring variable iterations.
• Comparison of weight matrix calculation methods identified a superior Bayesian-based approach.
• The method successfully detected novel conserved motifs.
• The procedure demonstrated robustness with adjustable false positive inclusion rates.

Conclusions:

• The described iterative approach is effective for analyzing protein sequence databases and discovering conserved segments.
• The Bayesian-based weight matrix calculation method enhances motif detection sensitivity.
• This technique offers a valuable tool for identifying functionally or structurally important motifs in uncharacterized proteins.