The threading approach to the inverse protein folding problem

Conference · Mon Dec 01 04:00:00 EST 1997

OSTI ID:549026

Smith, T F; Bienkowska, J; Rogers, B ^[1]

---

1. Boston Univ., MA (United States); and others

The logic behind the threading approach to the prediction of an amino acid sequence`s expected three-dimensional fold is almost seductively simple, if not obvious. In practice, however, the threading approach has not yet lived up to our expectations. Before asking why, let`s review what is generally meant by this approach to protein fold prediction. Threading rests on two basic ideas: first, that there is a limited and rather small number of basic {open_quotes}protein domain{close_quotes} core folds or architectures that need to be modeled; and, second, that a sum over a sequence`s amino acid structural environment preferences is a sufficient indicator for the recognition of native-like folds. The first of these ideas arises from our understanding of polymer chemistry, which suggests that there is only a limited number of ways to fold a repetitive polymer of two basic unit types (hydrophilic and hydrophobic), and from the observation that the vast majority of determined structures fall into only a few core architectures. The latter was perhaps first clearly stated by Jane Richardson in her {open_quotes}taxonomy of proteins{close_quotes}. First, protein structures and even their functions are known to be very robust to amino acid substitutions, an obvious requirement of any successfully evolving system. As has been noted many times, the many hundreds of distinct globin amino acid sequences fold to nearly identical structures, thus supporting the idea that some average over a sequence`s structural environmental preferences is as important as the atomic level details. Based on these two combined ideas, it has seemed reasonable, given a library of all expected basic fold architectures and an understanding of which average properties of amino acids determine their structural environment preferences, that we should be able to recognize which basic fold is preferred by which sequence. 17 refs., 5 figs.

• Research Organization: Association for Computing Machinery, New York, NY (United States); Sloan (Alfred P.) Foundation, New York, NY (United States)
• OSTI ID: 549026
• Report Number(s): CONF-970137--; CNN: Grant P41 LM02205-13
• Country of Publication: United States
• Language: English

Related Subjects

55 BIOLOGY AND MEDICINE
BASIC STUDIES
99 GENERAL AND MISCELLANEOUS
ALGORITHMS
AMINO ACID SEQUENCE
AMINO ACIDS
BIOLOGICAL EVOLUTION
CHEMISTRY
DNA SEQUENCING
HYDROPHYLIC POLYMERS
MATHEMATICAL MODELS
MOLECULAR BIOLOGY
PROTEIN STRUCTURE
PROTEINS
STRUCTURAL MODELS
STRUCTURE-ACTIVITY RELATIONSHIPS
TAXONOMY
THREE-DIMENSIONAL CALCULATIONS