Related Experiment Videos
Patterns, structures, and amino acid frequencies in structural building blocks, a protein secondary structure
J S Fetrow1, M J Palumbo, G Berg
1Department of Biological Sciences, University at Albany, SUNY 12222, USA. acque@isadora.albany.edu
Proteins
|February 1, 1997
Summary
Researchers developed a method using artificial neural networks to classify protein local structures into Structural Building Blocks (SBBs). This approach aids in understanding protein structure and sequence relationships for improved structure prediction.
Area of Science:
- * Computational Biology
- * Structural Bioinformatics
- * Artificial Intelligence in Biology
Background:
- * Understanding local protein structures is crucial for deciphering protein function and dynamics.
- * Previous work introduced an autoassociative artificial neural network (autoANN) for discovering intrinsic macromolecular structural features.
- * The hidden unit activations from autoANN provide a low-dimensional encoding of local protein backbone structure.
Purpose of the Study:
- * To apply and validate an autoANN-based method for classifying protein local structural features, termed Structural Building Blocks (SBBs).
- * To analyze amino acid frequencies and common patterns within identified SBBs.
- * To explore the utility of SBBs as building blocks for protein structure analysis and prediction.
Main Methods:
- * Application of a previously developed autoassociative artificial neural network (autoANN) to a large protein database.
- * Clustering of autoANN hidden unit activations to define Structural Building Blocks (SBBs).
- * Analysis of amino acid preferences and SBB patterns in protein local structures.
Main Results:
- * The SBB classification method successfully identifies regular secondary structures (alpha helix, beta strand) and other local features like helix and strand caps.
- * Significant amino acid preferences were observed at specific positions within certain SBBs.
- * Distinct patterns of SBBs were identified in protein regions previously described as 'random coil', revealing underlying structural motifs.
- * The method demonstrates applicability for protein structure classification.
Conclusions:
- * Structural Building Blocks (SBBs) provide a novel classification of local protein structural features.
- * SBB analysis reveals sequence-structure relationships and identifies structural motifs in protein backbones.
- * This method offers a valuable tool for enhancing the understanding of protein sequence-structure relationships and improving protein structure prediction.