Machine discovery of protein motifs

Abstract

Abstract. The investigation of relations between protein tertiary structure and amino acid sequence is a topic of tremendous importance in molecular biology. The automated discovery of recurrent patterns of structure and sequence is an essential part of this investigation. These patterns, known as protein motifs, are abstractions of fragments drawn from proteins of known sequence and tertiary structure. This paper has two objectives. The first is to introduce and define protein motifs, and provide a survey of previous research on protein motif discovery. The second is to present and apply a novel approach to protein motif representation and discovery, which is based on a spatial description logic and the symbolic machine learning paradigm of structured concept formation. A large database of protein fragments is processed using this approach, and several interesting and significant protein motifs are discovered.

Untitled

“…The following section proposes a representation that has a model-theoretic semantics for both structure and sequence motifs and addresses many of the issues raised in this survey. Taylor, 1986Smith & Smith, 1990Rooman & Wodak, 1988Rooman & Wodak, 1991Rooman et al, 1990bHunter & States, 1991Rooman et al, 1990aNussinov & Wolfson, 1991Prestrelski et al, 1992Onizuka et al, 1993Matsuo & Kanehisa, 1993Unger et al, 1989Blundell et al, 1987Sali & Blundell, 1990Zhang et al, 1993Conklin et al, 1993 …”

“…Both approaches require a knowledge representation, such as SVC, capable of describing multilevel structured objects. Initial results on level-2 protein motif discovery with IMEM is reported by Conklin et al (1994). Future work could also focus on using discovered heptamer motifs to parse or "tile" a protein.…”

“…An earlier paper (Conklin et al, 1993) showed how SVC could be used to represent all four protein motif types. The discussion here will be mainly concerned with structuresequence motifs.…”

Structure Comparison and Structure Patterns

“…This can also be extended to include structure descriptions, so that if pattern P matches structure S, then S ¿ P (or S 2 L.P /). This is easiest to do formally for deterministic patterns (see, e.g., Shinohara and Arikawa (1995) and Conklin (1995)). For probabilistic patterns, it is not that straightforward, but for example for alignments, a subalignment (consecutive columns) can be seen as a generalization of an alignment.…”

A novel approach to knowledge discovery and representation in biological databases

“…In the context of biological sequence analysis, a motif is a region or portion of protein or DNA/RNA sequences that has specific structure and significant function. Protein motifs have been divided into four categories: sequence motifs, sequence-structure motifs, structure motifs and structure-sequence motifs [13]. Sequence motifs are linear strings of amino acid residues with an implicit topological ordering and they are the most commonly encountered motif type in molecular biology.…”